id sid tid token lemma pos cana-3719 1 1 communications communication NOUN cana-3719 1 2 on on ADP cana-3719 1 3 applied apply VERB cana-3719 1 4 nonlinear nonlinear ADJ cana-3719 1 5 analysis analysis NOUN cana-3719 1 6 issn issn NOUN cana-3719 1 7 : : PUNCT cana-3719 1 8 1074 1074 NUM cana-3719 1 9 - - PUNCT cana-3719 1 10 133x 133x NUM cana-3719 1 11 vol vol NOUN cana-3719 1 12 32 32 NUM cana-3719 1 13 no no NOUN cana-3719 1 14 . . PUNCT cana-3719 2 1 8s 8s PROPN cana-3719 2 2 ( ( PUNCT cana-3719 2 3 2025 2025 NUM cana-3719 2 4 ) ) PUNCT cana-3719 2 5 577 577 NUM cana-3719 2 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 2 7 advances advance NOUN cana-3719 2 8 in in ADP cana-3719 2 9 deep deep ADJ cana-3719 2 10 learning learn VERB cana-3719 2 11 image image NOUN cana-3719 2 12 inpainting inpainte VERB cana-3719 2 13 architectures architecture NOUN cana-3719 2 14 for for ADP cana-3719 2 15 the the DET cana-3719 2 16 effective effective ADJ cana-3719 2 17 reconstruction reconstruction NOUN cana-3719 2 18 of of ADP cana-3719 2 19 damaged damage VERB cana-3719 2 20 regionsa regionsa ADJ cana-3719 2 21 systematic systematic ADJ cana-3719 2 22 review review PROPN cana-3719 2 23 * * PROPN cana-3719 2 24 suma suma PROPN cana-3719 2 25 n1 n1 PROPN cana-3719 2 26 , , PUNCT cana-3719 3 1 dr dr PROPN cana-3719 3 2 . . PROPN cana-3719 4 1 kusuma kusuma PROPN cana-3719 4 2 kumari kumari PROPN cana-3719 4 3 b.m2 b.m2 PRON cana-3719 4 4 1research 1research NUM cana-3719 4 5 scholar scholar NOUN cana-3719 4 6 , , PUNCT cana-3719 4 7 department department NOUN cana-3719 4 8 of of ADP cana-3719 4 9 studies study NOUN cana-3719 4 10 and and CCONJ cana-3719 4 11 research research NOUN cana-3719 4 12 in in ADP cana-3719 4 13 computer computer NOUN cana-3719 4 14 applications application NOUN cana-3719 4 15 , , PUNCT cana-3719 4 16 jnanasiri jnanasiri PROPN cana-3719 4 17 campus campus PROPN cana-3719 4 18 , , PUNCT cana-3719 4 19 tumkur tumkur PROPN cana-3719 4 20 university university PROPN cana-3719 4 21 , , PUNCT cana-3719 4 22 bidrakatte bidrakatte NOUN cana-3719 4 23 , , PUNCT cana-3719 4 24 tumkuru tumkuru PROPN cana-3719 4 25 , , PUNCT cana-3719 4 26 karnataka karnataka PROPN cana-3719 4 27 , , PUNCT cana-3719 4 28 india india PROPN cana-3719 4 29 2assistant 2assistant PROPN cana-3719 4 30 professor professor NOUN cana-3719 4 31 , , PUNCT cana-3719 4 32 department department NOUN cana-3719 4 33 of of ADP cana-3719 4 34 studies study NOUN cana-3719 4 35 and and CCONJ cana-3719 4 36 research research NOUN cana-3719 4 37 in in ADP cana-3719 4 38 computer computer NOUN cana-3719 4 39 applications application NOUN cana-3719 4 40 , , PUNCT cana-3719 4 41 jnanasiri jnanasiri PROPN cana-3719 4 42 campus campus PROPN cana-3719 4 43 , , PUNCT cana-3719 4 44 tumkur tumkur PROPN cana-3719 4 45 university university PROPN cana-3719 4 46 , , PUNCT cana-3719 4 47 bidrakatte bidrakatte PROPN cana-3719 4 48 , , PUNCT cana-3719 4 49 tumkuru tumkuru PROPN cana-3719 4 50 , , PUNCT cana-3719 4 51 karnataka karnataka PROPN cana-3719 4 52 , , PUNCT cana-3719 4 53 india india PROPN cana-3719 4 54 corresponding corresponding PROPN cana-3719 4 55 author author NOUN cana-3719 4 56 : : PUNCT cana-3719 4 57 suma suma PROPN cana-3719 4 58 n n PRON cana-3719 4 59 article article NOUN cana-3719 4 60 history history NOUN cana-3719 4 61 : : PUNCT cana-3719 4 62 received receive VERB cana-3719 4 63 : : PUNCT cana-3719 4 64 06 06 NUM cana-3719 4 65 - - SYM cana-3719 4 66 11 11 NUM cana-3719 4 67 - - PUNCT cana-3719 4 68 2024 2024 NUM cana-3719 4 69 revised revise VERB cana-3719 4 70 : : PUNCT cana-3719 4 71 13 13 NUM cana-3719 4 72 - - SYM cana-3719 4 73 12 12 NUM cana-3719 4 74 - - PUNCT cana-3719 4 75 2024 2024 NUM cana-3719 4 76 accepted accept VERB cana-3719 4 77 : : PUNCT cana-3719 4 78 02 02 NUM cana-3719 4 79 - - PUNCT cana-3719 4 80 01 01 NUM cana-3719 4 81 - - PUNCT cana-3719 4 82 2025 2025 NUM cana-3719 4 83 abstract abstract NOUN cana-3719 4 84 : : PUNCT cana-3719 4 85 image image NOUN cana-3719 4 86 inpainting inpainte VERB cana-3719 4 87 is be AUX cana-3719 4 88 the the DET cana-3719 4 89 method method NOUN cana-3719 4 90 of of ADP cana-3719 4 91 reconstructing reconstruct VERB cana-3719 4 92 the the DET cana-3719 4 93 missed miss VERB cana-3719 4 94 or or CCONJ cana-3719 4 95 damaged damaged ADJ cana-3719 4 96 region region NOUN cana-3719 4 97 of of ADP cana-3719 4 98 the the DET cana-3719 4 99 picture picture NOUN cana-3719 4 100 with with ADP cana-3719 4 101 certain certain ADJ cana-3719 4 102 rules rule NOUN cana-3719 4 103 that that PRON cana-3719 4 104 play play VERB cana-3719 4 105 a a DET cana-3719 4 106 vital vital ADJ cana-3719 4 107 role role NOUN cana-3719 4 108 in in ADP cana-3719 4 109 computer computer NOUN cana-3719 4 110 vision vision NOUN cana-3719 4 111 applications application NOUN cana-3719 4 112 . . PUNCT cana-3719 5 1 numerous numerous ADJ cana-3719 5 2 image image NOUN cana-3719 5 3 inpainting inpainte VERB cana-3719 5 4 approaches approach NOUN cana-3719 5 5 have have AUX cana-3719 5 6 been be AUX cana-3719 5 7 recently recently ADV cana-3719 5 8 proposed propose VERB cana-3719 5 9 that that SCONJ cana-3719 5 10 effectively effectively ADV cana-3719 5 11 reconstruct reconstruct VERB cana-3719 5 12 the the DET cana-3719 5 13 missed miss VERB cana-3719 5 14 regions region NOUN cana-3719 5 15 . . PUNCT cana-3719 6 1 traditionally traditionally ADV cana-3719 6 2 the the DET cana-3719 6 3 missed miss VERB cana-3719 6 4 regions region NOUN cana-3719 6 5 are be AUX cana-3719 6 6 restored restore VERB cana-3719 6 7 by by ADP cana-3719 6 8 schemes scheme NOUN cana-3719 6 9 that that PRON cana-3719 6 10 are be AUX cana-3719 6 11 derived derive VERB cana-3719 6 12 from from ADP cana-3719 6 13 the the DET cana-3719 6 14 diffusion diffusion NOUN cana-3719 6 15 models model NOUN cana-3719 6 16 , , PUNCT cana-3719 6 17 exemplar exemplar ADJ cana-3719 6 18 approaches approach NOUN cana-3719 6 19 , , PUNCT cana-3719 6 20 and and CCONJ cana-3719 6 21 sparsity sparsity NOUN cana-3719 6 22 approaches approach NOUN cana-3719 6 23 . . PUNCT cana-3719 7 1 different different ADJ cana-3719 7 2 hybrid hybrid ADJ cana-3719 7 3 schemes scheme NOUN cana-3719 7 4 are be AUX cana-3719 7 5 derived derive VERB cana-3719 7 6 from from ADP cana-3719 7 7 these these DET cana-3719 7 8 traditional traditional ADJ cana-3719 7 9 schemes scheme NOUN cana-3719 7 10 that that PRON cana-3719 7 11 show show VERB cana-3719 7 12 reasonable reasonable ADJ cana-3719 7 13 performance performance NOUN cana-3719 7 14 in in ADP cana-3719 7 15 missed missed ADJ cana-3719 7 16 region region NOUN cana-3719 7 17 reconstruction reconstruction NOUN cana-3719 7 18 . . PUNCT cana-3719 8 1 recently recently ADV cana-3719 8 2 deep deep ADV cana-3719 8 3 learning learn VERB cana-3719 8 4 -based -based ADJ cana-3719 8 5 approaches approach NOUN cana-3719 8 6 which which PRON cana-3719 8 7 are be AUX cana-3719 8 8 derived derive VERB cana-3719 8 9 from from ADP cana-3719 8 10 the the DET cana-3719 8 11 architectures architecture NOUN cana-3719 8 12 of of ADP cana-3719 8 13 generative generative ADJ cana-3719 8 14 adversarial adversarial ADJ cana-3719 8 15 networks network NOUN cana-3719 8 16 ( ( PUNCT cana-3719 8 17 gan gan PROPN cana-3719 8 18 ) ) PUNCT cana-3719 8 19 , , PUNCT cana-3719 8 20 convolutional convolutional ADJ cana-3719 8 21 neural neural ADJ cana-3719 8 22 networks network NOUN cana-3719 8 23 ( ( PUNCT cana-3719 8 24 cnn cnn PROPN cana-3719 8 25 ) ) PUNCT cana-3719 8 26 , , PUNCT cana-3719 8 27 transformers transformer NOUN cana-3719 8 28 , , PUNCT cana-3719 8 29 and and CCONJ cana-3719 8 30 u u NOUN cana-3719 8 31 - - NOUN cana-3719 8 32 net net NOUN cana-3719 8 33 are be AUX cana-3719 8 34 commonly commonly ADV cana-3719 8 35 used use VERB cana-3719 8 36 for for ADP cana-3719 8 37 effective effective ADJ cana-3719 8 38 image image NOUN cana-3719 8 39 painting painting NOUN cana-3719 8 40 . . PUNCT cana-3719 9 1 more more ADV cana-3719 9 2 specifically specifically ADV cana-3719 9 3 numerous numerous ADJ cana-3719 9 4 approaches approach NOUN cana-3719 9 5 have have AUX cana-3719 9 6 been be AUX cana-3719 9 7 derived derive VERB cana-3719 9 8 using use VERB cana-3719 9 9 the the DET cana-3719 9 10 advantage advantage NOUN cana-3719 9 11 of of ADP cana-3719 9 12 gan gan PROPN cana-3719 9 13 and and CCONJ cana-3719 9 14 cnn cnn PROPN cana-3719 9 15 architectures architecture NOUN cana-3719 9 16 . . PUNCT cana-3719 10 1 these these DET cana-3719 10 2 schemes scheme NOUN cana-3719 10 3 focus focus VERB cana-3719 10 4 on on ADP cana-3719 10 5 the the DET cana-3719 10 6 reconstruction reconstruction NOUN cana-3719 10 7 of of ADP cana-3719 10 8 global global ADJ cana-3719 10 9 structures structure NOUN cana-3719 10 10 and and CCONJ cana-3719 10 11 local local ADJ cana-3719 10 12 texture texture ADJ cana-3719 10 13 components component NOUN cana-3719 10 14 of of ADP cana-3719 10 15 the the DET cana-3719 10 16 missed miss VERB cana-3719 10 17 region region NOUN cana-3719 10 18 from from ADP cana-3719 10 19 the the DET cana-3719 10 20 known know VERB cana-3719 10 21 region region NOUN cana-3719 10 22 . . PUNCT cana-3719 11 1 therefore therefore ADV cana-3719 11 2 , , PUNCT cana-3719 11 3 the the DET cana-3719 11 4 paper paper NOUN cana-3719 11 5 provides provide VERB cana-3719 11 6 a a DET cana-3719 11 7 review review NOUN cana-3719 11 8 of of ADP cana-3719 11 9 different different ADJ cana-3719 11 10 traditional traditional ADJ cana-3719 11 11 , , PUNCT cana-3719 11 12 and and CCONJ cana-3719 11 13 recent recent ADJ cana-3719 11 14 deep deep ADJ cana-3719 11 15 -learning -learning ADJ cana-3719 11 16 schemes scheme NOUN cana-3719 11 17 . . PUNCT cana-3719 12 1 keywords keyword NOUN cana-3719 12 2 : : PUNCT cana-3719 12 3 image image NOUN cana-3719 12 4 inpainting inpainting NOUN cana-3719 12 5 , , PUNCT cana-3719 12 6 exemplar exemplar ADJ cana-3719 12 7 approach approach NOUN cana-3719 12 8 , , PUNCT cana-3719 12 9 gan gan PROPN cana-3719 12 10 approach approach NOUN cana-3719 12 11 , , PUNCT cana-3719 12 12 convolutional convolutional ADJ cana-3719 12 13 neural neural ADJ cana-3719 12 14 network network NOUN cana-3719 12 15 , , PUNCT cana-3719 12 16 gan gan PROPN cana-3719 12 17 model model NOUN cana-3719 12 18 . . PUNCT cana-3719 13 1 1 1 X cana-3719 13 2 . . X cana-3719 13 3 introduction introduction NOUN cana-3719 13 4 the the DET cana-3719 13 5 art art NOUN cana-3719 13 6 of of ADP cana-3719 13 7 restoring restore VERB cana-3719 13 8 damaged damage VERB cana-3719 13 9 photographs photograph NOUN cana-3719 13 10 or or CCONJ cana-3719 13 11 paintings painting NOUN cana-3719 13 12 that that PRON cana-3719 13 13 have have VERB cana-3719 13 14 minor minor ADJ cana-3719 13 15 damages damage NOUN cana-3719 13 16 namely namely ADV cana-3719 13 17 spots spot NOUN cana-3719 13 18 , , PUNCT cana-3719 13 19 dusts dust NOUN cana-3719 13 20 , , PUNCT cana-3719 13 21 cracks crack NOUN cana-3719 13 22 , , PUNCT cana-3719 13 23 and and CCONJ cana-3719 13 24 scratches scratch NOUN cana-3719 13 25 is be AUX cana-3719 13 26 termed term VERB cana-3719 13 27 to to PART cana-3719 13 28 be be AUX cana-3719 13 29 image image NOUN cana-3719 13 30 inpainting inpainte VERB cana-3719 13 31 [ [ X cana-3719 13 32 1 1 NUM cana-3719 13 33 ] ] PUNCT cana-3719 13 34 . . PUNCT cana-3719 14 1 the the DET cana-3719 14 2 image image NOUN cana-3719 14 3 inpainting inpainte VERB cana-3719 14 4 procedure procedure NOUN cana-3719 14 5 improves improve VERB cana-3719 14 6 the the DET cana-3719 14 7 visual visual ADJ cana-3719 14 8 appearance appearance NOUN cana-3719 14 9 of of ADP cana-3719 14 10 the the DET cana-3719 14 11 images image NOUN cana-3719 14 12 . . PUNCT cana-3719 15 1 the the DET cana-3719 15 2 digital digital ADJ cana-3719 15 3 image image NOUN cana-3719 15 4 inpainting inpainte VERB cana-3719 15 5 usually usually ADV cana-3719 15 6 interpolates interpolate VERB cana-3719 15 7 the the DET cana-3719 15 8 neighborhood neighborhood NOUN cana-3719 15 9 of of ADP cana-3719 15 10 the the DET cana-3719 15 11 missed miss VERB cana-3719 15 12 region region NOUN cana-3719 15 13 to to PART cana-3719 15 14 restore restore VERB cana-3719 15 15 the the DET cana-3719 15 16 missed miss VERB cana-3719 15 17 area area NOUN cana-3719 15 18 . . PUNCT cana-3719 16 1 i.e. i.e. X cana-3719 16 2 the the DET cana-3719 16 3 known know VERB cana-3719 16 4 picture picture NOUN cana-3719 16 5 information information NOUN cana-3719 16 6 is be AUX cana-3719 16 7 utilized utilize VERB cana-3719 16 8 to to PART cana-3719 16 9 estimate estimate VERB cana-3719 16 10 the the DET cana-3719 16 11 damaged damage VERB cana-3719 16 12 region region NOUN cana-3719 16 13 . . PUNCT cana-3719 17 1 image image NOUN cana-3719 17 2 inpainting inpainte VERB cana-3719 17 3 [ [ X cana-3719 17 4 2 2 NUM cana-3719 17 5 ] ] PUNCT cana-3719 17 6 has have VERB cana-3719 17 7 different different ADJ cana-3719 17 8 applications application NOUN cana-3719 17 9 namely namely ADV cana-3719 17 10 old old ADJ cana-3719 17 11 photo photo NOUN cana-3719 17 12 coloring coloring NOUN cana-3719 17 13 , , PUNCT cana-3719 17 14 scratch scratch NOUN cana-3719 17 15 removal removal NOUN cana-3719 17 16 , , PUNCT cana-3719 17 17 text text NOUN cana-3719 17 18 removal removal NOUN cana-3719 17 19 , , PUNCT cana-3719 17 20 watermark watermark NOUN cana-3719 17 21 removal removal NOUN cana-3719 17 22 , , PUNCT cana-3719 17 23 object object VERB cana-3719 17 24 / / SYM cana-3719 17 25 undesired undesired ADJ cana-3719 17 26 region region NOUN cana-3719 17 27 removal removal NOUN cana-3719 17 28 , , PUNCT cana-3719 17 29 etc etc X cana-3719 17 30 . . X cana-3719 18 1 the the DET cana-3719 18 2 different different ADJ cana-3719 18 3 applications application NOUN cana-3719 18 4 of of ADP cana-3719 18 5 image image NOUN cana-3719 18 6 inpainting inpainte VERB cana-3719 18 7 are be AUX cana-3719 18 8 illustrated illustrate VERB cana-3719 18 9 in in ADP cana-3719 18 10 fig fig NOUN cana-3719 18 11 . . PUNCT cana-3719 19 1 1 1 X cana-3719 19 2 . . X cana-3719 19 3 such such ADJ cana-3719 19 4 applications application NOUN cana-3719 19 5 can can AUX cana-3719 19 6 be be AUX cana-3719 19 7 broadly broadly ADV cana-3719 19 8 classified classify VERB cana-3719 19 9 as as ADP cana-3719 19 10 restoration restoration NOUN cana-3719 19 11 of of ADP cana-3719 19 12 damaged damage VERB cana-3719 19 13 regions region NOUN cana-3719 19 14 , , PUNCT cana-3719 19 15 object object NOUN cana-3719 19 16 removal removal NOUN cana-3719 19 17 , , PUNCT cana-3719 19 18 image image NOUN cana-3719 19 19 completion completion NOUN cana-3719 19 20 , , PUNCT cana-3719 19 21 and and CCONJ cana-3719 19 22 artistic artistic ADJ cana-3719 19 23 applications application NOUN cana-3719 19 24 [ [ X cana-3719 19 25 3 3 NUM cana-3719 19 26 ] ] PUNCT cana-3719 19 27 . . PUNCT cana-3719 20 1 fig fig NOUN cana-3719 20 2 . . PUNCT cana-3719 21 1 1 1 NUM cana-3719 21 2 : : PUNCT cana-3719 21 3 practical practical ADJ cana-3719 21 4 use use NOUN cana-3719 21 5 of of ADP cana-3719 21 6 image image NOUN cana-3719 21 7 inpainting inpainte VERB cana-3719 21 8 algorithms algorithms NOUN cana-3719 21 9 communications communication NOUN cana-3719 21 10 on on ADP cana-3719 21 11 applied apply VERB cana-3719 21 12 nonlinear nonlinear ADJ cana-3719 21 13 analysis analysis NOUN cana-3719 21 14 issn issn NOUN cana-3719 21 15 : : PUNCT cana-3719 21 16 1074 1074 NUM cana-3719 21 17 - - PUNCT cana-3719 21 18 133x 133x NUM cana-3719 21 19 vol vol NOUN cana-3719 21 20 32 32 NUM cana-3719 21 21 no no NOUN cana-3719 21 22 . . PUNCT cana-3719 22 1 8s 8s PROPN cana-3719 22 2 ( ( PUNCT cana-3719 22 3 2025 2025 NUM cana-3719 22 4 ) ) PUNCT cana-3719 22 5 578 578 NUM cana-3719 22 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 22 7 the the DET cana-3719 22 8 image image NOUN cana-3719 22 9 inpainting inpainte VERB cana-3719 22 10 process process NOUN cana-3719 22 11 should should AUX cana-3719 22 12 preserve preserve VERB cana-3719 22 13 the the DET cana-3719 22 14 semantic semantic ADJ cana-3719 22 15 structure structure NOUN cana-3719 22 16 while while SCONJ cana-3719 22 17 synthesizing synthesize VERB cana-3719 22 18 the the DET cana-3719 22 19 damaged damage VERB cana-3719 22 20 or or CCONJ cana-3719 22 21 unknown unknown ADJ cana-3719 22 22 region region NOUN cana-3719 22 23 . . PUNCT cana-3719 23 1 the the DET cana-3719 23 2 image image NOUN cana-3719 23 3 inpainting inpainte VERB cana-3719 23 4 also also ADV cana-3719 23 5 improves improve VERB cana-3719 23 6 the the DET cana-3719 23 7 resolution resolution NOUN cana-3719 23 8 and and CCONJ cana-3719 23 9 helps help VERB cana-3719 23 10 to to PART cana-3719 23 11 recover recover VERB cana-3719 23 12 the the DET cana-3719 23 13 images image NOUN cana-3719 23 14 that that PRON cana-3719 23 15 are be AUX cana-3719 23 16 damaged damage VERB cana-3719 23 17 due due ADP cana-3719 23 18 to to ADP cana-3719 23 19 storage storage NOUN cana-3719 23 20 , , PUNCT cana-3719 23 21 and and CCONJ cana-3719 23 22 transmission transmission NOUN cana-3719 23 23 . . PUNCT cana-3719 24 1 the the DET cana-3719 24 2 two two NUM cana-3719 24 3 broad broad ADJ cana-3719 24 4 classes class NOUN cana-3719 24 5 of of ADP cana-3719 24 6 image image NOUN cana-3719 24 7 inpainting inpainte VERB cana-3719 24 8 schemes scheme NOUN cana-3719 24 9 are be AUX cana-3719 24 10 the the DET cana-3719 24 11 deep deep ADJ cana-3719 24 12 learning learning NOUN cana-3719 24 13 approach approach NOUN cana-3719 24 14 and and CCONJ cana-3719 24 15 the the DET cana-3719 24 16 traditional traditional ADJ cana-3719 24 17 approach approach NOUN cana-3719 24 18 as as SCONJ cana-3719 24 19 illustrated illustrate VERB cana-3719 24 20 in in ADP cana-3719 24 21 fig fig NOUN cana-3719 24 22 . . PUNCT cana-3719 25 1 2 2 X cana-3719 25 2 . . X cana-3719 25 3 the the DET cana-3719 25 4 traditional traditional ADJ cana-3719 25 5 schemes scheme NOUN cana-3719 25 6 include include VERB cana-3719 25 7 diffusion diffusion NOUN cana-3719 25 8 - - PUNCT cana-3719 25 9 based base VERB cana-3719 25 10 schemes scheme NOUN cana-3719 25 11 [ [ X cana-3719 25 12 4 4 NUM cana-3719 25 13 ] ] PUNCT cana-3719 25 14 , , PUNCT cana-3719 25 15 exemplar exemplar NOUN cana-3719 25 16 - - PUNCT cana-3719 25 17 based base VERB cana-3719 25 18 schemes scheme NOUN cana-3719 25 19 [ [ X cana-3719 25 20 5 5 NUM cana-3719 25 21 ] ] PUNCT cana-3719 25 22 , , PUNCT cana-3719 25 23 and and CCONJ cana-3719 25 24 sparsitybased sparsitybase VERB cana-3719 25 25 schemes scheme NOUN cana-3719 25 26 [ [ X cana-3719 25 27 6 6 NUM cana-3719 25 28 ] ] PUNCT cana-3719 25 29 . . PUNCT cana-3719 26 1 fig fig NOUN cana-3719 26 2 . . PUNCT cana-3719 27 1 2 2 NUM cana-3719 27 2 : : PUNCT cana-3719 27 3 classification classification NOUN cana-3719 27 4 of of ADP cana-3719 27 5 various various ADJ cana-3719 27 6 image image NOUN cana-3719 27 7 inpainting inpainte VERB cana-3719 27 8 schemes scheme NOUN cana-3719 27 9 different different ADJ cana-3719 27 10 deep deep ADJ cana-3719 27 11 learning learning NOUN cana-3719 27 12 architectures architecture NOUN cana-3719 27 13 are be AUX cana-3719 27 14 derived derive VERB cana-3719 27 15 from from ADP cana-3719 27 16 the the DET cana-3719 27 17 cnn cnn NOUN cana-3719 28 1 [ [ X cana-3719 28 2 7 7 NUM cana-3719 28 3 ] ] PUNCT cana-3719 28 4 , , PUNCT cana-3719 28 5 gan gan X cana-3719 28 6 [ [ X cana-3719 28 7 8 8 NUM cana-3719 28 8 ] ] PUNCT cana-3719 28 9 , , PUNCT cana-3719 28 10 transformer transformer NOUN cana-3719 28 11 [ [ X cana-3719 28 12 9 9 NUM cana-3719 28 13 ] ] PUNCT cana-3719 28 14 , , PUNCT cana-3719 28 15 and and CCONJ cana-3719 28 16 u u ADJ cana-3719 28 17 - - ADJ cana-3719 28 18 net net ADJ cana-3719 28 19 structures structure NOUN cana-3719 28 20 [ [ X cana-3719 28 21 10 10 NUM cana-3719 28 22 ] ] PUNCT cana-3719 28 23 . . PUNCT cana-3719 29 1 there there PRON cana-3719 29 2 are be VERB cana-3719 29 3 two two NUM cana-3719 29 4 challenges challenge NOUN cana-3719 29 5 in in ADP cana-3719 29 6 image image NOUN cana-3719 29 7 inpainting inpainte VERB cana-3719 29 8 application application NOUN cana-3719 29 9 from from ADP cana-3719 29 10 a a DET cana-3719 29 11 practical practical ADJ cana-3719 29 12 perspective perspective NOUN cana-3719 29 13 . . PUNCT cana-3719 30 1 ( ( PUNCT cana-3719 30 2 i i NOUN cana-3719 30 3 ) ) PUNCT cana-3719 30 4 restoring restore VERB cana-3719 30 5 the the DET cana-3719 30 6 missed miss VERB cana-3719 30 7 or or CCONJ cana-3719 30 8 damaged damaged ADJ cana-3719 30 9 region region NOUN cana-3719 30 10 from from ADP cana-3719 30 11 the the DET cana-3719 30 12 known know VERB cana-3719 30 13 region region NOUN cana-3719 30 14 to to PART cana-3719 30 15 obtain obtain VERB cana-3719 30 16 the the DET cana-3719 30 17 complete complete ADJ cana-3719 30 18 images image NOUN cana-3719 30 19 . . PUNCT cana-3719 31 1 such such ADJ cana-3719 31 2 challenges challenge NOUN cana-3719 31 3 are be AUX cana-3719 31 4 present present ADJ cana-3719 31 5 in in ADP cana-3719 31 6 the the DET cana-3719 31 7 reconstruction reconstruction NOUN cana-3719 31 8 of of ADP cana-3719 31 9 old old ADJ cana-3719 31 10 photographs photograph NOUN cana-3719 31 11 . . PUNCT cana-3719 32 1 ( ( PUNCT cana-3719 32 2 ii ii NOUN cana-3719 32 3 ) ) PUNCT cana-3719 32 4 removal removal NOUN cana-3719 32 5 of of ADP cana-3719 32 6 undesirable undesirable ADJ cana-3719 32 7 content content NOUN cana-3719 32 8 from from ADP cana-3719 32 9 the the DET cana-3719 32 10 picture picture NOUN cana-3719 32 11 , , PUNCT cana-3719 32 12 and and CCONJ cana-3719 32 13 replacing replace VERB cana-3719 32 14 the the DET cana-3719 32 15 region region NOUN cana-3719 32 16 with with ADP cana-3719 32 17 plausible plausible ADJ cana-3719 32 18 content content NOUN cana-3719 32 19 . . PUNCT cana-3719 33 1 such such DET cana-3719 33 2 a a DET cana-3719 33 3 challenge challenge NOUN cana-3719 33 4 occurs occur VERB cana-3719 33 5 mostly mostly ADV cana-3719 33 6 in in ADP cana-3719 33 7 photo photo NOUN cana-3719 33 8 editing edit VERB cana-3719 33 9 for for ADP cana-3719 33 10 removing remove VERB cana-3719 33 11 the the DET cana-3719 33 12 objects object NOUN cana-3719 33 13 and and CCONJ cana-3719 33 14 the the DET cana-3719 33 15 watermark watermark NOUN cana-3719 33 16 content content NOUN cana-3719 33 17 . . PUNCT cana-3719 34 1 also also ADV cana-3719 34 2 , , PUNCT cana-3719 34 3 the the DET cana-3719 34 4 behavior behavior NOUN cana-3719 34 5 of of ADP cana-3719 34 6 the the DET cana-3719 34 7 picture picture NOUN cana-3719 34 8 inpainting inpainte VERB cana-3719 34 9 scheme scheme NOUN cana-3719 34 10 depends depend VERB cana-3719 34 11 on on ADP cana-3719 34 12 the the DET cana-3719 34 13 mask mask NOUN cana-3719 34 14 ratio ratio NOUN cana-3719 34 15 . . PUNCT cana-3719 35 1 the the DET cana-3719 35 2 paper paper NOUN cana-3719 35 3 has have VERB cana-3719 35 4 the the DET cana-3719 35 5 following follow VERB cana-3719 35 6 forthcoming forthcoming ADJ cana-3719 35 7 sections section NOUN cana-3719 35 8 . . PUNCT cana-3719 36 1 section section NOUN cana-3719 36 2 2 2 NUM cana-3719 36 3 deliberates deliberate VERB cana-3719 36 4 the the DET cana-3719 36 5 outperforming outperform VERB cana-3719 36 6 traditional traditional ADJ cana-3719 36 7 inpainting inpainting ADJ cana-3719 36 8 schemes scheme NOUN cana-3719 36 9 and and CCONJ cana-3719 36 10 section section NOUN cana-3719 36 11 3 3 NUM cana-3719 36 12 discusses discuss VERB cana-3719 36 13 the the DET cana-3719 36 14 works work NOUN cana-3719 36 15 that that PRON cana-3719 36 16 are be AUX cana-3719 36 17 related relate VERB cana-3719 36 18 to to ADP cana-3719 36 19 deep deep ADJ cana-3719 36 20 learning learning NOUN cana-3719 36 21 . . PUNCT cana-3719 37 1 section section NOUN cana-3719 37 2 4 4 NUM cana-3719 37 3 discusses discuss VERB cana-3719 37 4 the the DET cana-3719 37 5 dataset dataset NOUN cana-3719 37 6 and and CCONJ cana-3719 37 7 important important ADJ cana-3719 37 8 evaluation evaluation NOUN cana-3719 37 9 measures measure NOUN cana-3719 37 10 used use VERB cana-3719 37 11 for for ADP cana-3719 37 12 the the DET cana-3719 37 13 analysis analysis NOUN cana-3719 37 14 of of ADP cana-3719 37 15 image image NOUN cana-3719 37 16 inpainting inpainte VERB cana-3719 37 17 schemes scheme NOUN cana-3719 37 18 . . PUNCT cana-3719 38 1 section section NOUN cana-3719 38 2 5 5 NUM cana-3719 38 3 shows show VERB cana-3719 38 4 the the DET cana-3719 38 5 performance performance NOUN cana-3719 38 6 comparison comparison NOUN cana-3719 38 7 of of ADP cana-3719 38 8 different different ADJ cana-3719 38 9 approaches approach NOUN cana-3719 38 10 and and CCONJ cana-3719 38 11 lastly lastly ADV cana-3719 38 12 , , PUNCT cana-3719 38 13 the the DET cana-3719 38 14 conclusion conclusion NOUN cana-3719 38 15 of of ADP cana-3719 38 16 the the DET cana-3719 38 17 paper paper NOUN cana-3719 38 18 is be AUX cana-3719 38 19 illustrated illustrate VERB cana-3719 38 20 in in ADP cana-3719 38 21 section section NOUN cana-3719 38 22 6 6 NUM cana-3719 38 23 . . NOUN cana-3719 38 24 2 2 NUM cana-3719 38 25 . . PUNCT cana-3719 38 26 traditional traditional ADJ cana-3719 38 27 schemes scheme NOUN cana-3719 38 28 the the DET cana-3719 38 29 traditional traditional ADJ cana-3719 38 30 image image NOUN cana-3719 38 31 inpainting inpainte VERB cana-3719 38 32 schemes scheme NOUN cana-3719 38 33 include include VERB cana-3719 38 34 the the DET cana-3719 38 35 approaches approach NOUN cana-3719 38 36 that that PRON cana-3719 38 37 are be AUX cana-3719 38 38 derived derive VERB cana-3719 38 39 from from ADP cana-3719 38 40 the the DET cana-3719 38 41 diffusion diffusion NOUN cana-3719 38 42 model model NOUN cana-3719 38 43 , , PUNCT cana-3719 38 44 exemplar exemplar ADJ cana-3719 38 45 approach approach NOUN cana-3719 38 46 , , PUNCT cana-3719 38 47 and and CCONJ cana-3719 38 48 sparsity sparsity NOUN cana-3719 38 49 approach approach NOUN cana-3719 38 50 . . PUNCT cana-3719 39 1 this this DET cana-3719 39 2 approach approach NOUN cana-3719 39 3 uses use VERB cana-3719 39 4 mathematical mathematical ADJ cana-3719 39 5 rules rule NOUN cana-3719 39 6 to to PART cana-3719 39 7 derive derive VERB cana-3719 39 8 the the DET cana-3719 39 9 missed miss VERB cana-3719 39 10 pixels pixel NOUN cana-3719 39 11 . . PUNCT cana-3719 40 1 2.1 2.1 NUM cana-3719 40 2 diffusion diffusion NOUN cana-3719 40 3 model model NOUN cana-3719 40 4 - - PUNCT cana-3719 40 5 based base VERB cana-3719 40 6 approach approach NOUN cana-3719 40 7 diffusion diffusion NOUN cana-3719 40 8 models model NOUN cana-3719 40 9 [ [ X cana-3719 40 10 11 11 NUM cana-3719 40 11 ] ] PUNCT cana-3719 40 12 are be AUX cana-3719 40 13 utilized utilize VERB cana-3719 40 14 in in ADP cana-3719 40 15 diffusion diffusion NOUN cana-3719 40 16 - - PUNCT cana-3719 40 17 based base VERB cana-3719 40 18 image image NOUN cana-3719 40 19 inpainting inpainte VERB cana-3719 40 20 . . PUNCT cana-3719 41 1 the the DET cana-3719 41 2 two two NUM cana-3719 41 3 processes process NOUN cana-3719 41 4 involved involve VERB cana-3719 41 5 in in ADP cana-3719 41 6 diffusion diffusion NOUN cana-3719 41 7 - - PUNCT cana-3719 41 8 based base VERB cana-3719 41 9 inpainting inpainting NOUN cana-3719 41 10 are be AUX cana-3719 41 11 reverse reverse ADJ cana-3719 41 12 and and CCONJ cana-3719 41 13 forward forward ADJ cana-3719 41 14 processes process NOUN cana-3719 41 15 . . PUNCT cana-3719 42 1 the the DET cana-3719 42 2 forward forward ADJ cana-3719 42 3 procedure procedure NOUN cana-3719 42 4 adds add VERB cana-3719 42 5 the the DET cana-3719 42 6 noise noise NOUN cana-3719 42 7 in in ADP cana-3719 42 8 multiple multiple ADJ cana-3719 42 9 steps step NOUN cana-3719 42 10 so so SCONJ cana-3719 42 11 that that SCONJ cana-3719 42 12 the the DET cana-3719 42 13 image image NOUN cana-3719 42 14 becomes become VERB cana-3719 42 15 completely completely ADV cana-3719 42 16 noisy noisy ADJ cana-3719 42 17 . . PUNCT cana-3719 43 1 the the DET cana-3719 43 2 stochastic stochastic ADJ cana-3719 43 3 - - PUNCT cana-3719 43 4 differential differential NOUN cana-3719 43 5 equation equation NOUN cana-3719 43 6 is be AUX cana-3719 43 7 used use VERB cana-3719 43 8 to to PART cana-3719 43 9 represent represent VERB cana-3719 43 10 the the DET cana-3719 43 11 forward forward ADJ cana-3719 43 12 process process NOUN cana-3719 43 13 . . PUNCT cana-3719 44 1 the the DET cana-3719 44 2 noise noise NOUN cana-3719 44 3 added add VERB cana-3719 44 4 an an DET cana-3719 44 5 image image NOUN cana-3719 44 6 at at ADP cana-3719 44 7 iteration iteration NOUN cana-3719 44 8 𝑖 𝑖 PRON cana-3719 44 9 can can AUX cana-3719 44 10 be be AUX cana-3719 44 11 represented represent VERB cana-3719 44 12 as as ADP cana-3719 44 13 , , PUNCT cana-3719 44 14 𝐼𝑖 𝐼𝑖 PROPN cana-3719 44 15 = = PUNCT cana-3719 44 16 √1 √1 PROPN cana-3719 45 1 + + CCONJ cana-3719 45 2 𝛾𝑖 𝛾𝑖 ADV cana-3719 45 3 πœŒπ‘– πœŒπ‘– ADP cana-3719 45 4 + + CCONJ cana-3719 45 5 βˆšπ›Ύπ‘– βˆšπ›Ύπ‘– PRON cana-3719 45 6 πΌπ‘œ πΌπ‘œ PROPN cana-3719 45 7 ( ( PUNCT cana-3719 45 8 1 1 NUM cana-3719 45 9 ) ) PUNCT cana-3719 45 10 here here ADV cana-3719 45 11 , , PUNCT cana-3719 45 12 πΌπ‘œ πΌπ‘œ PROPN cana-3719 45 13 resembles resemble VERB cana-3719 45 14 the the DET cana-3719 45 15 input input NOUN cana-3719 45 16 image image NOUN cana-3719 45 17 , , PUNCT cana-3719 45 18 𝛾𝑖 𝛾𝑖 CCONJ cana-3719 45 19 controls control VERB cana-3719 45 20 the the DET cana-3719 45 21 noise noise NOUN cana-3719 45 22 over over ADP cana-3719 45 23 iteration iteration NOUN cana-3719 45 24 𝑖 𝑖 NOUN cana-3719 45 25 , , PUNCT cana-3719 45 26 πœŒπ‘– πœŒπ‘– ADP cana-3719 45 27 represents represent VERB cana-3719 45 28 the the DET cana-3719 45 29 noise noise NOUN cana-3719 45 30 that that PRON cana-3719 45 31 is be AUX cana-3719 45 32 added add VERB cana-3719 45 33 at at ADP cana-3719 45 34 iteration iteration NOUN cana-3719 45 35 𝑖. 𝑖. ADV cana-3719 45 36 the the DET cana-3719 45 37 reverse reverse NOUN cana-3719 45 38 noise noise NOUN cana-3719 45 39 process process NOUN cana-3719 45 40 removes remove VERB cana-3719 45 41 the the DET cana-3719 45 42 noise noise NOUN cana-3719 45 43 to to PART cana-3719 45 44 reconstruct reconstruct VERB cana-3719 45 45 the the DET cana-3719 45 46 original original ADJ cana-3719 45 47 image image NOUN cana-3719 45 48 . . PUNCT cana-3719 46 1 this this PRON cana-3719 46 2 was be AUX cana-3719 46 3 also also ADV cana-3719 46 4 performed perform VERB cana-3719 46 5 iteratively iteratively ADV cana-3719 46 6 . . PUNCT cana-3719 47 1 the the DET cana-3719 47 2 two two NUM cana-3719 47 3 regions region NOUN cana-3719 47 4 used use VERB cana-3719 47 5 in in ADP cana-3719 47 6 image image NOUN cana-3719 47 7 inpainting inpainte VERB cana-3719 47 8 include include VERB cana-3719 47 9 the the DET cana-3719 47 10 unmasked unmasked ADJ cana-3719 47 11 region region NOUN cana-3719 47 12 and and CCONJ cana-3719 47 13 the the DET cana-3719 47 14 masked masked ADJ cana-3719 47 15 region region NOUN cana-3719 47 16 . . PUNCT cana-3719 48 1 the the DET cana-3719 48 2 visible visible ADJ cana-3719 48 3 part part NOUN cana-3719 48 4 of of ADP cana-3719 48 5 the the DET cana-3719 48 6 image image NOUN cana-3719 48 7 is be AUX cana-3719 48 8 termed term VERB cana-3719 48 9 be be AUX cana-3719 48 10 unmasked unmasked ADJ cana-3719 48 11 region region NOUN cana-3719 48 12 . . PUNCT cana-3719 49 1 the the DET cana-3719 49 2 unmasked unmasked ADJ cana-3719 49 3 region region NOUN cana-3719 49 4 helps help VERB cana-3719 49 5 to to PART cana-3719 49 6 identify identify VERB cana-3719 49 7 the the DET cana-3719 49 8 mask mask NOUN cana-3719 49 9 area area NOUN cana-3719 49 10 content content NOUN cana-3719 49 11 . . PUNCT cana-3719 50 1 the the DET cana-3719 50 2 mask mask NOUN cana-3719 50 3 area area NOUN cana-3719 50 4 resembles resemble VERB cana-3719 50 5 the the DET cana-3719 50 6 picture picture NOUN cana-3719 50 7 region region NOUN cana-3719 50 8 , , PUNCT cana-3719 50 9 where where SCONJ cana-3719 50 10 the the DET cana-3719 50 11 image image NOUN cana-3719 50 12 pixels pixel NOUN cana-3719 50 13 are be AUX cana-3719 50 14 corrupted corrupt VERB cana-3719 50 15 , , PUNCT cana-3719 50 16 missed miss VERB cana-3719 50 17 , , PUNCT cana-3719 50 18 or or CCONJ cana-3719 50 19 intentionally intentionally ADV cana-3719 50 20 modified modify VERB cana-3719 50 21 . . PUNCT cana-3719 51 1 it it PRON cana-3719 51 2 also also ADV cana-3719 51 3 may may AUX cana-3719 51 4 be be AUX cana-3719 51 5 due due ADJ cana-3719 51 6 to to ADP cana-3719 51 7 communications communication NOUN cana-3719 51 8 on on ADP cana-3719 51 9 applied apply VERB cana-3719 51 10 nonlinear nonlinear ADJ cana-3719 51 11 analysis analysis NOUN cana-3719 51 12 issn issn NOUN cana-3719 51 13 : : PUNCT cana-3719 51 14 1074 1074 NUM cana-3719 51 15 - - PUNCT cana-3719 51 16 133x 133x NUM cana-3719 51 17 vol vol NOUN cana-3719 51 18 32 32 NUM cana-3719 51 19 no no NOUN cana-3719 51 20 . . PUNCT cana-3719 52 1 8s 8s PROPN cana-3719 52 2 ( ( PUNCT cana-3719 52 3 2025 2025 NUM cana-3719 52 4 ) ) PUNCT cana-3719 52 5 579 579 NUM cana-3719 52 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 52 7 an an DET cana-3719 52 8 occlusion occlusion NOUN cana-3719 52 9 ( ( PUNCT cana-3719 52 10 unwanted unwanted ADJ cana-3719 52 11 objects object NOUN cana-3719 52 12 ) ) PUNCT cana-3719 52 13 . . PUNCT cana-3719 53 1 the the DET cana-3719 53 2 model model NOUN cana-3719 53 3 learns learn VERB cana-3719 53 4 to to PART cana-3719 53 5 recover recover VERB cana-3719 53 6 the the DET cana-3719 53 7 visual visual ADJ cana-3719 53 8 content content NOUN cana-3719 53 9 from from ADP cana-3719 53 10 the the DET cana-3719 53 11 noisy noisy ADJ cana-3719 53 12 image image NOUN cana-3719 53 13 on on ADP cana-3719 53 14 the the DET cana-3719 53 15 unmasked unmasked ADJ cana-3719 53 16 region region NOUN cana-3719 53 17 . . PUNCT cana-3719 54 1 the the DET cana-3719 54 2 same same ADJ cana-3719 54 3 recovery recovery NOUN cana-3719 54 4 procedure procedure NOUN cana-3719 54 5 is be AUX cana-3719 54 6 followed follow VERB cana-3719 54 7 in in ADP cana-3719 54 8 the the DET cana-3719 54 9 masked masked ADJ cana-3719 54 10 region region NOUN cana-3719 54 11 to to PART cana-3719 54 12 complete complete VERB cana-3719 54 13 the the DET cana-3719 54 14 image image NOUN cana-3719 54 15 inpainting inpainte VERB cana-3719 54 16 process process NOUN cana-3719 54 17 . . PUNCT cana-3719 55 1 the the DET cana-3719 55 2 forward forward ADJ cana-3719 55 3 process process NOUN cana-3719 55 4 depends depend VERB cana-3719 55 5 on on ADP cana-3719 55 6 parameters parameter NOUN cana-3719 55 7 such such ADJ cana-3719 55 8 as as ADP cana-3719 55 9 the the DET cana-3719 55 10 number number NOUN cana-3719 55 11 of of ADP cana-3719 55 12 diffusion diffusion NOUN cana-3719 55 13 steps step NOUN cana-3719 55 14 , , PUNCT cana-3719 55 15 noise noise NOUN cana-3719 55 16 type type NOUN cana-3719 55 17 , , PUNCT cana-3719 55 18 and and CCONJ cana-3719 55 19 noise noise NOUN cana-3719 55 20 schedule schedule NOUN cana-3719 55 21 . . PUNCT cana-3719 56 1 the the DET cana-3719 56 2 reverse reverse ADJ cana-3719 56 3 process process NOUN cana-3719 56 4 depends depend VERB cana-3719 56 5 on on ADP cana-3719 56 6 the the DET cana-3719 56 7 model model NOUN cana-3719 56 8 , , PUNCT cana-3719 56 9 noise noise NOUN cana-3719 56 10 , , PUNCT cana-3719 56 11 predictor predictor NOUN cana-3719 56 12 , , PUNCT cana-3719 56 13 and and CCONJ cana-3719 56 14 conditioning conditioning NOUN cana-3719 56 15 mechanism mechanism NOUN cana-3719 56 16 . . PUNCT cana-3719 57 1 diffusion diffusion NOUN cana-3719 57 2 schemes scheme NOUN cana-3719 57 3 namely namely ADV cana-3719 57 4 denoising denoise VERB cana-3719 57 5 diffusion diffusion NOUN cana-3719 57 6 implicit implicit ADJ cana-3719 57 7 models model NOUN cana-3719 57 8 ( ( PUNCT cana-3719 57 9 ddim ddim NOUN cana-3719 57 10 ) ) PUNCT cana-3719 58 1 [ [ X cana-3719 58 2 12 12 NUM cana-3719 58 3 ] ] PUNCT cana-3719 58 4 , , PUNCT cana-3719 58 5 and and CCONJ cana-3719 58 6 denoising denoise VERB cana-3719 58 7 diffusion diffusion NOUN cana-3719 58 8 probabilistic probabilistic ADJ cana-3719 58 9 models model NOUN cana-3719 58 10 ( ( PUNCT cana-3719 58 11 ddpm ddpm PROPN cana-3719 58 12 ) ) PUNCT cana-3719 59 1 [ [ X cana-3719 59 2 13 13 NUM cana-3719 59 3 ] ] PUNCT cana-3719 59 4 are be AUX cana-3719 59 5 commonly commonly ADV cana-3719 59 6 used use VERB cana-3719 59 7 diffusion diffusion NOUN cana-3719 59 8 models model NOUN cana-3719 59 9 . . PUNCT cana-3719 60 1 2.2 2.2 NUM cana-3719 60 2 exemplar exemplar NOUN cana-3719 60 3 - - PUNCT cana-3719 60 4 based base VERB cana-3719 60 5 schemes scheme NOUN cana-3719 60 6 the the DET cana-3719 60 7 exemplar exemplar ADJ cana-3719 60 8 approach approach NOUN cana-3719 60 9 [ [ X cana-3719 60 10 14 14 NUM cana-3719 60 11 ] ] PUNCT cana-3719 60 12 aims aim VERB cana-3719 60 13 to to PART cana-3719 60 14 minimize minimize VERB cana-3719 60 15 the the DET cana-3719 60 16 energy energy NOUN cana-3719 60 17 function function NOUN cana-3719 60 18 𝑃 𝑃 NOUN cana-3719 60 19 = = PUNCT cana-3719 60 20 𝛼𝑃𝑑(𝐼 𝛼𝑃𝑑(𝐼 NUM cana-3719 60 21 ) ) PUNCT cana-3719 60 22 + + X cana-3719 60 23 𝛽𝑃𝑠(𝐼 𝛽𝑃𝑠(𝐼 NOUN cana-3719 60 24 ) ) PUNCT cana-3719 60 25 ( ( PUNCT cana-3719 60 26 2 2 X cana-3719 60 27 ) ) PUNCT cana-3719 60 28 𝑃𝑑(𝐼 𝑃𝑑(𝐼 NOUN cana-3719 60 29 ) ) PUNCT cana-3719 60 30 resembles resemble VERB cana-3719 60 31 how how SCONJ cana-3719 60 32 the the DET cana-3719 60 33 inpainted inpainte VERB cana-3719 60 34 image image NOUN cana-3719 60 35 agrees agree VERB cana-3719 60 36 with with ADP cana-3719 60 37 the the DET cana-3719 60 38 known know VERB cana-3719 60 39 pixels pixel NOUN cana-3719 60 40 and and CCONJ cana-3719 60 41 𝑃𝑠(𝐼 𝑃𝑠(𝐼 ADV cana-3719 60 42 ) ) PUNCT cana-3719 60 43 resemble resemble VERB cana-3719 60 44 the the DET cana-3719 60 45 continuity continuity NOUN cana-3719 60 46 between between ADP cana-3719 60 47 the the DET cana-3719 60 48 inpainted inpainte VERB cana-3719 60 49 , , PUNCT cana-3719 60 50 and and CCONJ cana-3719 60 51 original original ADJ cana-3719 60 52 region region NOUN cana-3719 60 53 . . PUNCT cana-3719 61 1 𝛼 𝛼 X cana-3719 61 2 , , PUNCT cana-3719 61 3 and and CCONJ cana-3719 61 4 𝛽 𝛽 NOUN cana-3719 61 5 resembles resemble VERB cana-3719 61 6 the the DET cana-3719 61 7 weight weight NOUN cana-3719 61 8 to to PART cana-3719 61 9 control control VERB cana-3719 61 10 the the DET cana-3719 61 11 data datum NOUN cana-3719 61 12 term term NOUN cana-3719 61 13 , , PUNCT cana-3719 61 14 and and CCONJ cana-3719 61 15 smoothness smoothness ADJ cana-3719 61 16 term term NOUN cana-3719 61 17 respectively respectively ADV cana-3719 61 18 . . PUNCT cana-3719 62 1 the the DET cana-3719 62 2 exemplar exemplar ADJ cana-3719 62 3 approach approach NOUN cana-3719 62 4 uses use VERB cana-3719 62 5 the the DET cana-3719 62 6 principle principle NOUN cana-3719 62 7 of of ADP cana-3719 62 8 borrowing borrow VERB cana-3719 62 9 the the DET cana-3719 62 10 information information NOUN cana-3719 62 11 ( ( PUNCT cana-3719 62 12 known know VERB cana-3719 62 13 as as ADP cana-3719 62 14 exemplars exemplar NOUN cana-3719 62 15 ) ) PUNCT cana-3719 62 16 from from ADP cana-3719 62 17 the the DET cana-3719 62 18 content content NOUN cana-3719 62 19 available available ADJ cana-3719 62 20 region region NOUN cana-3719 62 21 to to PART cana-3719 62 22 estimate estimate VERB cana-3719 62 23 the the DET cana-3719 62 24 content content NOUN cana-3719 62 25 of of ADP cana-3719 62 26 the the DET cana-3719 62 27 masked masked ADJ cana-3719 62 28 area area NOUN cana-3719 62 29 . . PUNCT cana-3719 63 1 this this DET cana-3719 63 2 approach approach NOUN cana-3719 63 3 uses use VERB cana-3719 63 4 a a DET cana-3719 63 5 patch patch NOUN cana-3719 63 6 - - PUNCT cana-3719 63 7 based base VERB cana-3719 63 8 reconstruction reconstruction NOUN cana-3719 63 9 [ [ X cana-3719 63 10 15 15 NUM cana-3719 63 11 ] ] PUNCT cana-3719 63 12 instead instead ADV cana-3719 63 13 of of ADP cana-3719 63 14 using use VERB cana-3719 63 15 pixel pixel ADJ cana-3719 63 16 - - ADJ cana-3719 63 17 wise wise ADJ cana-3719 63 18 interpolation interpolation NOUN cana-3719 63 19 . . PUNCT cana-3719 64 1 the the DET cana-3719 64 2 main main ADJ cana-3719 64 3 process process NOUN cana-3719 64 4 involved involve VERB cana-3719 64 5 in in ADP cana-3719 64 6 this this DET cana-3719 64 7 exemplar exemplar ADJ cana-3719 64 8 approach approach NOUN cana-3719 64 9 is be AUX cana-3719 64 10 selecting select VERB cana-3719 64 11 the the DET cana-3719 64 12 patch patch NOUN cana-3719 64 13 from from ADP cana-3719 64 14 the the DET cana-3719 64 15 known know VERB cana-3719 64 16 region region NOUN cana-3719 64 17 ( ( PUNCT cana-3719 64 18 unmasked unmasked ADJ cana-3719 64 19 region region NOUN cana-3719 64 20 ) ) PUNCT cana-3719 64 21 . . PUNCT cana-3719 65 1 the the DET cana-3719 65 2 selected select VERB cana-3719 65 3 patch patch NOUN cana-3719 65 4 should should AUX cana-3719 65 5 closely closely ADV cana-3719 65 6 match match VERB cana-3719 65 7 the the DET cana-3719 65 8 neighborhood neighborhood NOUN cana-3719 65 9 of of ADP cana-3719 65 10 the the DET cana-3719 65 11 masked masked ADJ cana-3719 65 12 region region NOUN cana-3719 65 13 . . PUNCT cana-3719 66 1 the the DET cana-3719 66 2 matching matching NOUN cana-3719 66 3 is be AUX cana-3719 66 4 done do VERB cana-3719 66 5 based base VERB cana-3719 66 6 on on ADP cana-3719 66 7 criteria criterion NOUN cana-3719 66 8 such such ADJ cana-3719 66 9 as as ADP cana-3719 66 10 contextual contextual ADJ cana-3719 66 11 consistency consistency NOUN cana-3719 66 12 , , PUNCT cana-3719 66 13 texture texture ADJ cana-3719 66 14 similarity similarity NOUN cana-3719 66 15 , , PUNCT cana-3719 66 16 or or CCONJ cana-3719 66 17 color color NOUN cana-3719 66 18 similarity similarity NOUN cana-3719 67 1 [ [ X cana-3719 67 2 16 16 NUM cana-3719 67 3 ] ] PUNCT cana-3719 67 4 . . PUNCT cana-3719 68 1 during during ADP cana-3719 68 2 the the DET cana-3719 68 3 patch patch NOUN cana-3719 68 4 transfer transfer NOUN cana-3719 68 5 , , PUNCT cana-3719 68 6 the the DET cana-3719 68 7 patches patch NOUN cana-3719 68 8 are be AUX cana-3719 68 9 aligned align VERB cana-3719 68 10 at at ADP cana-3719 68 11 the the DET cana-3719 68 12 boundaries boundary NOUN cana-3719 68 13 . . PUNCT cana-3719 69 1 to to PART cana-3719 69 2 have have VERB cana-3719 69 3 a a DET cana-3719 69 4 smooth smooth ADJ cana-3719 69 5 transition transition NOUN cana-3719 69 6 between between ADP cana-3719 69 7 the the DET cana-3719 69 8 patched patched ADJ cana-3719 69 9 region region NOUN cana-3719 69 10 , , PUNCT cana-3719 69 11 and and CCONJ cana-3719 69 12 the the DET cana-3719 69 13 boundaries boundary NOUN cana-3719 69 14 , , PUNCT cana-3719 69 15 the the DET cana-3719 69 16 region region NOUN cana-3719 69 17 is be AUX cana-3719 69 18 blended blend VERB cana-3719 69 19 [ [ X cana-3719 69 20 17 17 NUM cana-3719 69 21 ] ] PUNCT cana-3719 69 22 . . PUNCT cana-3719 70 1 usually usually ADV cana-3719 70 2 , , PUNCT cana-3719 70 3 the the DET cana-3719 70 4 masked mask VERB cana-3719 70 5 regions region NOUN cana-3719 70 6 are be AUX cana-3719 70 7 filled fill VERB cana-3719 70 8 with with ADP cana-3719 70 9 patches patch NOUN cana-3719 70 10 step step NOUN cana-3719 70 11 by by ADP cana-3719 70 12 step step NOUN cana-3719 70 13 . . PUNCT cana-3719 71 1 the the DET cana-3719 71 2 performance performance NOUN cana-3719 71 3 of of ADP cana-3719 71 4 the the DET cana-3719 71 5 exemplar exemplar NOUN cana-3719 71 6 - - PUNCT cana-3719 71 7 based base VERB cana-3719 71 8 approach approach NOUN cana-3719 71 9 depends depend VERB cana-3719 71 10 on on ADP cana-3719 71 11 parameters parameter NOUN cana-3719 71 12 such such ADJ cana-3719 71 13 as as ADP cana-3719 71 14 search search NOUN cana-3719 71 15 window window NOUN cana-3719 71 16 size size NOUN cana-3719 71 17 , , PUNCT cana-3719 71 18 patch patch NOUN cana-3719 71 19 overlap overlap NOUN cana-3719 71 20 , , PUNCT cana-3719 71 21 patch patch NOUN cana-3719 71 22 size size NOUN cana-3719 71 23 , , PUNCT cana-3719 71 24 boundary boundary ADJ cana-3719 71 25 propagation propagation NOUN cana-3719 71 26 strategy strategy NOUN cana-3719 71 27 , , PUNCT cana-3719 71 28 matching matching NOUN cana-3719 71 29 criteria criterion NOUN cana-3719 71 30 , , PUNCT cana-3719 71 31 inpainting inpainte VERB cana-3719 71 32 iterations iteration NOUN cana-3719 71 33 , , PUNCT cana-3719 71 34 patch patch ADJ cana-3719 71 35 matching matching NOUN cana-3719 71 36 distance distance NOUN cana-3719 71 37 , , PUNCT cana-3719 71 38 and and CCONJ cana-3719 71 39 patch patch NOUN cana-3719 71 40 search search NOUN cana-3719 71 41 strategy strategy NOUN cana-3719 71 42 . . PUNCT cana-3719 72 1 the the DET cana-3719 72 2 greedy greedy ADJ cana-3719 72 3 search search NOUN cana-3719 72 4 algorithm algorithm NOUN cana-3719 72 5 [ [ X cana-3719 72 6 18 18 NUM cana-3719 72 7 ] ] PUNCT cana-3719 72 8 is be AUX cana-3719 72 9 commonly commonly ADV cana-3719 72 10 used use VERB cana-3719 72 11 to to PART cana-3719 72 12 search search VERB cana-3719 72 13 the the DET cana-3719 72 14 patches patch NOUN cana-3719 72 15 . . PUNCT cana-3719 73 1 2.3 2.3 NUM cana-3719 73 2 sparsity sparsity NOUN cana-3719 73 3 - - PUNCT cana-3719 73 4 based base VERB cana-3719 73 5 schemes scheme NOUN cana-3719 73 6 this this DET cana-3719 73 7 approach approach NOUN cana-3719 73 8 uses use VERB cana-3719 73 9 the the DET cana-3719 73 10 sparsity sparsity NOUN cana-3719 73 11 concept concept NOUN cana-3719 73 12 to to PART cana-3719 73 13 estimate estimate VERB cana-3719 73 14 the the DET cana-3719 73 15 missed miss VERB cana-3719 73 16 region region NOUN cana-3719 73 17 . . PUNCT cana-3719 74 1 the the DET cana-3719 74 2 principle principle NOUN cana-3719 74 3 of of ADP cana-3719 74 4 this this DET cana-3719 74 5 approach approach NOUN cana-3719 74 6 is be AUX cana-3719 74 7 that that SCONJ cana-3719 74 8 a a DET cana-3719 74 9 natural natural ADJ cana-3719 74 10 image image NOUN cana-3719 74 11 can can AUX cana-3719 74 12 be be AUX cana-3719 74 13 compressed compress VERB cana-3719 74 14 to to ADP cana-3719 74 15 a a DET cana-3719 74 16 few few ADJ cana-3719 74 17 coefficients coefficient NOUN cana-3719 74 18 . . PUNCT cana-3719 75 1 these these DET cana-3719 75 2 coefficients coefficient NOUN cana-3719 75 3 act act VERB cana-3719 75 4 as as ADP cana-3719 75 5 a a DET cana-3719 75 6 dictionary dictionary ADJ cana-3719 75 7 element element NOUN cana-3719 75 8 or or CCONJ cana-3719 75 9 basis basis NOUN cana-3719 75 10 function function NOUN cana-3719 75 11 . . PUNCT cana-3719 76 1 fourier fourier NOUN cana-3719 76 2 , , PUNCT cana-3719 76 3 wavelet wavelet NOUN cana-3719 76 4 , , PUNCT cana-3719 76 5 and and CCONJ cana-3719 76 6 dct dct PROPN cana-3719 76 7 transform transform NOUN cana-3719 76 8 are be AUX cana-3719 76 9 utilized utilize VERB cana-3719 76 10 to to PART cana-3719 76 11 estimate estimate VERB cana-3719 76 12 the the DET cana-3719 76 13 coefficients coefficient NOUN cana-3719 76 14 . . PUNCT cana-3719 77 1 the the DET cana-3719 77 2 performance performance NOUN cana-3719 77 3 of of ADP cana-3719 77 4 this this DET cana-3719 77 5 approach approach NOUN cana-3719 77 6 depends depend VERB cana-3719 77 7 on on ADP cana-3719 77 8 the the DET cana-3719 77 9 transform transform NOUN cana-3719 77 10 type type NOUN cana-3719 77 11 , , PUNCT cana-3719 77 12 stopping stop VERB cana-3719 77 13 criterion criterion NOUN cana-3719 77 14 , , PUNCT cana-3719 77 15 optimization optimization NOUN cana-3719 77 16 type type NOUN cana-3719 77 17 , , PUNCT cana-3719 77 18 and and CCONJ cana-3719 77 19 sparsity sparsity NOUN cana-3719 77 20 regularization regularization NOUN cana-3719 77 21 parameter parameter NOUN cana-3719 77 22 Ξ» Ξ» PROPN cana-3719 77 23 . . PUNCT cana-3719 78 1 the the DET cana-3719 78 2 reconstructed reconstructed ADJ cana-3719 78 3 image image NOUN cana-3719 78 4 in in ADP cana-3719 78 5 sparse sparse ADJ cana-3719 78 6 based base VERB cana-3719 78 7 approach approach NOUN cana-3719 78 8 can can AUX cana-3719 78 9 be be AUX cana-3719 78 10 represented represent VERB cana-3719 78 11 as as ADP cana-3719 78 12 , , PUNCT cana-3719 78 13 πΌπ‘Ÿ πΌπ‘Ÿ PROPN cana-3719 78 14 = = SYM cana-3719 78 15 π‘Žπ‘Ÿπ‘”π‘šπ‘–π‘› π‘Žπ‘Ÿπ‘”π‘šπ‘–π‘› PROPN cana-3719 78 16 𝑐 𝑐 PROPN cana-3719 78 17 { { PUNCT cana-3719 78 18 ‖𝑄(𝐢 ‖𝑄(𝐢 PROPN cana-3719 78 19 βˆ’ βˆ’ PROPN cana-3719 78 20 𝐢𝑝)β€–2 𝐢𝑝)β€–2 NUM cana-3719 78 21 2 2 NUM cana-3719 78 22 + + CCONJ cana-3719 78 23 πœ†β€–π‘π‘ πœ†β€–π‘π‘ ADJ cana-3719 78 24 β€–1 β€–1 NOUN cana-3719 78 25 } } PUNCT cana-3719 78 26 ( ( PUNCT cana-3719 78 27 3 3 X cana-3719 78 28 ) ) PUNCT cana-3719 78 29 𝑄 𝑄 PRON cana-3719 78 30 represents represent VERB cana-3719 78 31 the the DET cana-3719 78 32 projection projection NOUN cana-3719 78 33 operator operator NOUN cana-3719 78 34 . . PUNCT cana-3719 79 1 𝐢𝑝 𝐢𝑝 PROPN cana-3719 79 2 and and CCONJ cana-3719 79 3 πœ† πœ† NOUN cana-3719 79 4 represents represent VERB cana-3719 79 5 the the DET cana-3719 79 6 partially partially ADV cana-3719 79 7 observed observe VERB cana-3719 79 8 image image NOUN cana-3719 79 9 , , PUNCT cana-3719 79 10 and and CCONJ cana-3719 79 11 regularization regularization NOUN cana-3719 79 12 factor factor NOUN cana-3719 79 13 respectively respectively ADV cana-3719 79 14 . . PUNCT cana-3719 80 1 ‖𝑍𝑐 ‖𝑍𝑐 PROPN cana-3719 80 2 β€–1 β€–1 PROPN cana-3719 80 3 resembles resemble VERB cana-3719 80 4 the the DET cana-3719 80 5 𝑙𝑖 𝑙𝑖 NOUN cana-3719 80 6 βˆ’ βˆ’ PROPN cana-3719 80 7 π‘›π‘œπ‘Ÿπ‘š π‘›π‘œπ‘Ÿπ‘š NOUN cana-3719 80 8 of of ADP cana-3719 80 9 the the DET cana-3719 80 10 sparse sparse ADJ cana-3719 80 11 coefficients coefficient NOUN cana-3719 80 12 , , PUNCT cana-3719 80 13 and and CCONJ cana-3719 80 14 𝐢 𝐢 PROPN cana-3719 80 15 resembles resemble VERB cana-3719 80 16 the the DET cana-3719 80 17 input input NOUN cana-3719 80 18 image image NOUN cana-3719 80 19 . . PUNCT cana-3719 81 1 the the DET cana-3719 81 2 sparsity sparsity NOUN cana-3719 81 3 approach approach NOUN cana-3719 81 4 [ [ X cana-3719 81 5 19 19 NUM cana-3719 81 6 ] ] X cana-3719 81 7 uses use VERB cana-3719 81 8 controllable controllable ADJ cana-3719 81 9 generation generation NOUN cana-3719 81 10 and and CCONJ cana-3719 81 11 content content NOUN cana-3719 81 12 initialization initialization NOUN cana-3719 81 13 to to PART cana-3719 81 14 perform perform VERB cana-3719 81 15 the the DET cana-3719 81 16 image image NOUN cana-3719 81 17 inpainting inpainte VERB cana-3719 81 18 . . PUNCT cana-3719 82 1 a a DET cana-3719 82 2 multi multi ADJ cana-3719 82 3 - - ADJ cana-3719 82 4 pixel pixel ADJ cana-3719 82 5 window window NOUN cana-3719 82 6 mechanism mechanism NOUN cana-3719 82 7 was be AUX cana-3719 82 8 also also ADV cana-3719 82 9 used use VERB cana-3719 82 10 [ [ PUNCT cana-3719 82 11 20 20 NUM cana-3719 82 12 ] ] PUNCT cana-3719 82 13 to to PART cana-3719 82 14 capture capture VERB cana-3719 82 15 more more ADV cana-3719 82 16 fine fine ADJ cana-3719 82 17 details detail NOUN cana-3719 82 18 . . PUNCT cana-3719 83 1 3 3 X cana-3719 83 2 . . X cana-3719 83 3 deep deep ADJ cana-3719 83 4 learning learning NOUN cana-3719 83 5 - - PUNCT cana-3719 83 6 based base VERB cana-3719 83 7 image image NOUN cana-3719 83 8 inpainting inpainte VERB cana-3719 83 9 several several ADJ cana-3719 83 10 researchers researcher NOUN cana-3719 83 11 developed develop VERB cana-3719 83 12 image image NOUN cana-3719 83 13 inpainting inpainte VERB cana-3719 83 14 schemes scheme NOUN cana-3719 83 15 that that PRON cana-3719 83 16 are be AUX cana-3719 83 17 derived derive VERB cana-3719 83 18 from from ADP cana-3719 83 19 , , PUNCT cana-3719 83 20 cnn cnn PROPN cana-3719 83 21 , , PUNCT cana-3719 83 22 gan gan PROPN cana-3719 83 23 , , PUNCT cana-3719 83 24 transformer transformer NOUN cana-3719 83 25 , , PUNCT cana-3719 83 26 u u NOUN cana-3719 83 27 - - NOUN cana-3719 83 28 net net ADJ cana-3719 83 29 , , PUNCT cana-3719 83 30 attention attention NOUN cana-3719 83 31 , , PUNCT cana-3719 83 32 and and CCONJ cana-3719 83 33 texture texture ADJ cana-3719 83 34 - - PUNCT cana-3719 83 35 structure structure NOUN cana-3719 83 36 models model NOUN cana-3719 83 37 . . PUNCT cana-3719 84 1 communications communication NOUN cana-3719 84 2 on on ADP cana-3719 84 3 applied apply VERB cana-3719 84 4 nonlinear nonlinear ADJ cana-3719 84 5 analysis analysis NOUN cana-3719 84 6 issn issn NOUN cana-3719 84 7 : : PUNCT cana-3719 84 8 1074 1074 NUM cana-3719 84 9 - - PUNCT cana-3719 84 10 133x 133x NUM cana-3719 84 11 vol vol NOUN cana-3719 84 12 32 32 NUM cana-3719 84 13 no no NOUN cana-3719 84 14 . . PUNCT cana-3719 85 1 8s 8s PROPN cana-3719 85 2 ( ( PUNCT cana-3719 85 3 2025 2025 NUM cana-3719 85 4 ) ) PUNCT cana-3719 85 5 580 580 NUM cana-3719 85 6 https://internationalpubls.com https://internationalpubls.com NUM cana-3719 85 7 3.1 3.1 NUM cana-3719 85 8 cnn cnn PROPN cana-3719 85 9 - - PUNCT cana-3719 85 10 based base VERB cana-3719 85 11 schemes scheme NOUN cana-3719 85 12 the the DET cana-3719 85 13 cnn cnn PROPN cana-3719 85 14 - - PUNCT cana-3719 85 15 based base VERB cana-3719 85 16 approach approach NOUN cana-3719 85 17 [ [ X cana-3719 85 18 21 21 NUM cana-3719 85 19 ] ] PUNCT cana-3719 85 20 usually usually ADV cana-3719 85 21 has have VERB cana-3719 85 22 an an DET cana-3719 85 23 encoder encoder NOUN cana-3719 85 24 , , PUNCT cana-3719 85 25 and and CCONJ cana-3719 85 26 decoder decoder NOUN cana-3719 85 27 unit unit NOUN cana-3719 85 28 where where SCONJ cana-3719 85 29 the the DET cana-3719 85 30 encoder encoder NOUN cana-3719 85 31 structure structure NOUN cana-3719 85 32 collects collect VERB cana-3719 85 33 the the DET cana-3719 85 34 high high ADJ cana-3719 85 35 - - PUNCT cana-3719 85 36 level level NOUN cana-3719 85 37 descriptors descriptor NOUN cana-3719 85 38 from from ADP cana-3719 85 39 the the DET cana-3719 85 40 visible visible ADJ cana-3719 85 41 region region NOUN cana-3719 85 42 sequence sequence NOUN cana-3719 85 43 of of ADP cana-3719 85 44 convolutional convolutional ADJ cana-3719 85 45 filters filter NOUN cana-3719 85 46 that that PRON cana-3719 85 47 increase increase VERB cana-3719 85 48 the the DET cana-3719 85 49 feature feature NOUN cana-3719 85 50 maps map NOUN cana-3719 85 51 . . PUNCT cana-3719 86 1 a a DET cana-3719 86 2 bottleneck bottleneck NOUN cana-3719 86 3 layer layer NOUN cana-3719 86 4 is be AUX cana-3719 86 5 used use VERB cana-3719 86 6 after after ADP cana-3719 86 7 the the DET cana-3719 86 8 encoder encoder NOUN cana-3719 86 9 which which PRON cana-3719 86 10 shows show VERB cana-3719 86 11 the the DET cana-3719 86 12 compressed compress VERB cana-3719 86 13 representation representation NOUN cana-3719 86 14 of of ADP cana-3719 86 15 the the DET cana-3719 86 16 feature feature NOUN cana-3719 86 17 map map NOUN cana-3719 86 18 . . PUNCT cana-3719 87 1 quaternion quaternion VERB cana-3719 87 2 cnn cnn PROPN cana-3719 88 1 [ [ X cana-3719 88 2 22 22 NUM cana-3719 88 3 ] ] PUNCT cana-3719 88 4 was be AUX cana-3719 88 5 proposed propose VERB cana-3719 88 6 by by ADP cana-3719 88 7 miao miao PROPN cana-3719 88 8 et et PROPN cana-3719 88 9 al al PROPN cana-3719 88 10 . . PUNCT cana-3719 89 1 that that PRON cana-3719 89 2 overcome overcome VERB cana-3719 89 3 the the DET cana-3719 89 4 two two NUM cana-3719 89 5 important important ADJ cana-3719 89 6 challenges challenge NOUN cana-3719 89 7 in in ADP cana-3719 89 8 quaternion quaternion ADJ cana-3719 89 9 matrix matrix NOUN cana-3719 89 10 estimation estimation NOUN cana-3719 89 11 . . PUNCT cana-3719 90 1 firstly firstly ADV cana-3719 90 2 the the DET cana-3719 90 3 performance performance NOUN cana-3719 90 4 of of ADP cana-3719 90 5 image image NOUN cana-3719 90 6 inpainting inpainte VERB cana-3719 90 7 for for ADP cana-3719 90 8 the the DET cana-3719 90 9 quaternion quaternion NOUN cana-3719 90 10 matrix matrix NOUN cana-3719 90 11 approach approach NOUN cana-3719 90 12 highly highly ADV cana-3719 90 13 depends depend VERB cana-3719 90 14 on on ADP cana-3719 90 15 the the DET cana-3719 90 16 type type NOUN cana-3719 90 17 of of ADP cana-3719 90 18 regularizer regularizer NOUN cana-3719 90 19 . . PUNCT cana-3719 91 1 a a DET cana-3719 91 2 regularizer regularizer NOUN cana-3719 91 3 wo will AUX cana-3719 91 4 n't not PART cana-3719 91 5 show show VERB cana-3719 91 6 a a DET cana-3719 91 7 better well ADJ cana-3719 91 8 performance performance NOUN cana-3719 91 9 for for ADP cana-3719 91 10 all all PRON cana-3719 91 11 - - PUNCT cana-3719 91 12 natural natural ADJ cana-3719 91 13 images image NOUN cana-3719 91 14 . . PUNCT cana-3719 92 1 secondly secondly ADV cana-3719 92 2 , , PUNCT cana-3719 92 3 due due ADP cana-3719 92 4 to to ADP cana-3719 92 5 the the DET cana-3719 92 6 non non ADJ cana-3719 92 7 - - ADJ cana-3719 92 8 commutativity commutativity ADJ cana-3719 92 9 property property NOUN cana-3719 92 10 in in ADP cana-3719 92 11 multiplication multiplication NOUN cana-3719 92 12 for for ADP cana-3719 92 13 the the DET cana-3719 92 14 quaternion quaternion NOUN cana-3719 92 15 matrix matrix NOUN cana-3719 92 16 , , PUNCT cana-3719 92 17 optimizing optimize VERB cana-3719 92 18 the the DET cana-3719 92 19 quaternion quaternion NOUN cana-3719 92 20 approach approach NOUN cana-3719 92 21 is be AUX cana-3719 92 22 difficult difficult ADJ cana-3719 92 23 . . PUNCT cana-3719 93 1 the the DET cana-3719 93 2 two two NUM cana-3719 93 3 challenges challenge NOUN cana-3719 93 4 are be AUX cana-3719 93 5 overcome overcome VERB cana-3719 93 6 by by ADP cana-3719 93 7 the the DET cana-3719 93 8 use use NOUN cana-3719 93 9 of of ADP cana-3719 93 10 untrained untrained ADJ cana-3719 93 11 quaternion quaternion PROPN cana-3719 93 12 cnn cnn PROPN cana-3719 94 1 [ [ X cana-3719 94 2 23 23 NUM cana-3719 94 3 ] ] PUNCT cana-3719 94 4 . . PUNCT cana-3719 95 1 the the DET cana-3719 95 2 loss loss NOUN cana-3719 95 3 function function NOUN cana-3719 95 4 in in ADP cana-3719 95 5 the the DET cana-3719 95 6 cnn cnn PROPN cana-3719 95 7 - - PUNCT cana-3719 95 8 based base VERB cana-3719 95 9 approach approach NOUN cana-3719 95 10 can can AUX cana-3719 95 11 be be AUX cana-3719 95 12 expressed express VERB cana-3719 95 13 as as ADP cana-3719 95 14 , , PUNCT cana-3719 95 15 ƞ(πΌπ‘Ÿ ƞ(πΌπ‘Ÿ X cana-3719 95 16 , , PUNCT cana-3719 95 17 𝐼𝑖 𝐼𝑖 PROPN cana-3719 95 18 ) ) PUNCT cana-3719 95 19 = = SYM cana-3719 95 20 πœ†Ζž πœ†Ζž PROPN cana-3719 96 1 π‘Ÿ π‘Ÿ NOUN cana-3719 96 2 ( ( PUNCT cana-3719 96 3 πΌπ‘Ÿ πΌπ‘Ÿ PROPN cana-3719 96 4 ) ) PUNCT cana-3719 96 5 + + CCONJ cana-3719 96 6 ƞ ƞ X cana-3719 96 7 𝑑 𝑑 NOUN cana-3719 96 8 ( ( PUNCT cana-3719 96 9 πΌπ‘Ÿ πΌπ‘Ÿ PROPN cana-3719 96 10 , , PUNCT cana-3719 96 11 𝐼𝑖 𝐼𝑖 PROPN cana-3719 96 12 ) ) PUNCT cana-3719 96 13 ( ( PUNCT cana-3719 96 14 4 4 X cana-3719 96 15 ) ) PUNCT cana-3719 96 16 ƞ ƞ NOUN cana-3719 96 17 π‘Ÿ π‘Ÿ NOUN cana-3719 96 18 ( ( PUNCT cana-3719 96 19 πΌπ‘Ÿ πΌπ‘Ÿ PROPN cana-3719 96 20 ) ) PUNCT cana-3719 96 21 resembles resemble VERB cana-3719 96 22 the the DET cana-3719 96 23 regularization regularization NOUN cana-3719 96 24 term term NOUN cana-3719 96 25 , , PUNCT cana-3719 96 26 whose whose DET cana-3719 96 27 weight weight NOUN cana-3719 96 28 is be AUX cana-3719 96 29 controlled control VERB cana-3719 96 30 by by ADP cana-3719 96 31 the the DET cana-3719 96 32 hyperparameter hyperparameter NOUN cana-3719 96 33 πœ†. πœ†. VERB cana-3719 97 1 ƞ ƞ PROPN cana-3719 97 2 𝑑 𝑑 PROPN cana-3719 97 3 ( ( PUNCT cana-3719 97 4 πΌπ‘Ÿ πΌπ‘Ÿ PROPN cana-3719 97 5 , , PUNCT cana-3719 97 6 𝐼𝑖 𝐼𝑖 PROPN cana-3719 97 7 ) ) PUNCT cana-3719 97 8 resembles resemble VERB cana-3719 97 9 the the DET cana-3719 97 10 data data NOUN cana-3719 97 11 loss loss NOUN cana-3719 97 12 between between ADP cana-3719 97 13 the the DET cana-3719 97 14 inpainted inpainte VERB cana-3719 97 15 picture picture NOUN cana-3719 97 16 πΌπ‘Ÿ πΌπ‘Ÿ ADP cana-3719 97 17 , , PUNCT cana-3719 97 18 and and CCONJ cana-3719 97 19 ground ground NOUN cana-3719 97 20 truth truth NOUN cana-3719 97 21 picture picture NOUN cana-3719 97 22 𝐼𝑖. 𝐼𝑖. PROPN cana-3719 97 23 an an DET cana-3719 97 24 adaptive adaptive ADJ cana-3719 97 25 fusion fusion NOUN cana-3719 97 26 process process NOUN cana-3719 97 27 [ [ X cana-3719 97 28 24 24 NUM cana-3719 97 29 ] ] PUNCT cana-3719 97 30 was be AUX cana-3719 97 31 introduced introduce VERB cana-3719 97 32 by by ADP cana-3719 97 33 zhu zhu PROPN cana-3719 97 34 et et PROPN cana-3719 97 35 al al PROPN cana-3719 97 36 . . PUNCT cana-3719 98 1 that that PRON cana-3719 98 2 transfers transfer VERB cana-3719 98 3 visual visual ADJ cana-3719 98 4 style style NOUN cana-3719 98 5 descriptors descriptor NOUN cana-3719 98 6 of of ADP cana-3719 98 7 known know VERB cana-3719 98 8 regions region NOUN cana-3719 98 9 to to ADP cana-3719 98 10 unknown unknown ADJ cana-3719 98 11 regions region NOUN cana-3719 98 12 . . PUNCT cana-3719 99 1 the the DET cana-3719 99 2 transformer transformer NOUN cana-3719 99 3 - - PUNCT cana-3719 99 4 based base VERB cana-3719 99 5 approach approach NOUN cana-3719 99 6 provides provide VERB cana-3719 99 7 a a DET cana-3719 99 8 better well ADJ cana-3719 99 9 global global ADJ cana-3719 99 10 correlation correlation NOUN cana-3719 99 11 , , PUNCT cana-3719 99 12 while while SCONJ cana-3719 99 13 the the DET cana-3719 99 14 cnn cnn PROPN cana-3719 99 15 - - PUNCT cana-3719 99 16 based base VERB cana-3719 99 17 approach approach NOUN cana-3719 99 18 can can AUX cana-3719 99 19 effectively effectively ADV cana-3719 99 20 reconstruct reconstruct VERB cana-3719 99 21 the the DET cana-3719 99 22 local local ADJ cana-3719 99 23 patterns pattern NOUN cana-3719 99 24 . . PUNCT cana-3719 100 1 the the DET cana-3719 100 2 global global ADJ cana-3719 100 3 , , PUNCT cana-3719 100 4 and and CCONJ cana-3719 100 5 local local ADJ cana-3719 100 6 advantages advantage NOUN cana-3719 100 7 of of ADP cana-3719 100 8 the the DET cana-3719 100 9 transformer transformer NOUN cana-3719 100 10 , , PUNCT cana-3719 100 11 and and CCONJ cana-3719 100 12 cnn cnn PROPN cana-3719 100 13 are be AUX cana-3719 100 14 combined combine VERB cana-3719 100 15 to to PART cana-3719 100 16 derive derive VERB cana-3719 100 17 a a DET cana-3719 100 18 local local ADJ cana-3719 100 19 - - PUNCT cana-3719 100 20 global global ADJ cana-3719 100 21 mixture mixture NOUN cana-3719 100 22 [ [ X cana-3719 100 23 25 25 NUM cana-3719 100 24 ] ] PUNCT cana-3719 100 25 . . PUNCT cana-3719 101 1 inception inception ADJ cana-3719 101 2 networks network NOUN cana-3719 101 3 [ [ X cana-3719 101 4 26 26 NUM cana-3719 101 5 ] ] PUNCT cana-3719 101 6 , , PUNCT cana-3719 101 7 [ [ X cana-3719 101 8 27 27 NUM cana-3719 101 9 ] ] PUNCT cana-3719 101 10 , , PUNCT cana-3719 101 11 [ [ X cana-3719 101 12 28 28 NUM cana-3719 101 13 ] ] X cana-3719 101 14 architectures architecture NOUN cana-3719 101 15 , , PUNCT cana-3719 101 16 resnet resnet VERB cana-3719 101 17 [ [ X cana-3719 101 18 29 29 NUM cana-3719 101 19 ] ] PUNCT cana-3719 101 20 , , PUNCT cana-3719 101 21 and and CCONJ cana-3719 101 22 vgg vgg NOUN cana-3719 102 1 [ [ X cana-3719 102 2 30 30 NUM cana-3719 102 3 ] ] PUNCT cana-3719 102 4 , , PUNCT cana-3719 102 5 structures structure NOUN cana-3719 102 6 are be AUX cana-3719 102 7 derived derive VERB cana-3719 102 8 from from ADP cana-3719 102 9 the the DET cana-3719 102 10 traditional traditional ADJ cana-3719 102 11 cnn cnn PROPN cana-3719 102 12 approach approach NOUN cana-3719 102 13 . . PUNCT cana-3719 103 1 3.2 3.2 NUM cana-3719 103 2 gan gin VERB cana-3719 103 3 - - PUNCT cana-3719 103 4 based base VERB cana-3719 103 5 schemes scheme NOUN cana-3719 103 6 fig fig NOUN cana-3719 103 7 . . PUNCT cana-3719 104 1 3 3 NUM cana-3719 104 2 : : PUNCT cana-3719 104 3 gan gin VERB cana-3719 104 4 - - PUNCT cana-3719 104 5 based base VERB cana-3719 104 6 architecture architecture NOUN cana-3719 104 7 in in ADP cana-3719 104 8 image image NOUN cana-3719 104 9 inpainting inpainte VERB cana-3719 104 10 the the DET cana-3719 104 11 two two NUM cana-3719 104 12 major major ADJ cana-3719 104 13 components component NOUN cana-3719 104 14 to to PART cana-3719 104 15 perform perform VERB cana-3719 104 16 image image NOUN cana-3719 104 17 inpainting inpainte VERB cana-3719 104 18 [ [ X cana-3719 104 19 31 31 NUM cana-3719 104 20 ] ] PUNCT cana-3719 104 21 in in ADP cana-3719 104 22 gan gan ADJ cana-3719 104 23 architecture architecture NOUN cana-3719 104 24 are be AUX cana-3719 104 25 the the DET cana-3719 104 26 generator generator NOUN cana-3719 104 27 module module NOUN cana-3719 104 28 and and CCONJ cana-3719 104 29 discriminator discriminator NOUN cana-3719 104 30 module module NOUN cana-3719 104 31 as as SCONJ cana-3719 104 32 depicted depict VERB cana-3719 104 33 in in ADP cana-3719 104 34 fig fig NOUN cana-3719 104 35 . . PUNCT cana-3719 105 1 3 3 X cana-3719 105 2 . . X cana-3719 106 1 the the DET cana-3719 106 2 total total ADJ cana-3719 106 3 objective objective ADJ cana-3719 106 4 function function NOUN cana-3719 106 5 in in ADP cana-3719 106 6 gan gan PROPN cana-3719 106 7 is be AUX cana-3719 106 8 πœ‚π‘‘ πœ‚π‘‘ ADP cana-3719 106 9 = = PUNCT cana-3719 106 10 πœ†πœ‚π‘π‘–π‘₯ πœ†πœ‚π‘π‘–π‘₯ NOUN cana-3719 106 11 + + CCONJ cana-3719 106 12 πœ‚πΊ πœ‚πΊ ADJ cana-3719 106 13 ( ( PUNCT cana-3719 106 14 5 5 NUM cana-3719 106 15 ) ) PUNCT cana-3719 106 16 here here ADV cana-3719 106 17 , , PUNCT cana-3719 106 18 πœ‚π‘π‘–π‘₯ πœ‚π‘π‘–π‘₯ NOUN cana-3719 106 19 represents represent VERB cana-3719 106 20 the the DET cana-3719 106 21 pixel pixel ADJ cana-3719 106 22 - - ADJ cana-3719 106 23 wise wise ADJ cana-3719 106 24 loss loss NOUN cana-3719 106 25 , , PUNCT cana-3719 106 26 and and CCONJ cana-3719 106 27 πœ‚πΊ πœ‚πΊ NOUN cana-3719 106 28 represents represent VERB cana-3719 106 29 the the DET cana-3719 106 30 generator generator NOUN cana-3719 106 31 loss loss NOUN cana-3719 106 32 . . PUNCT cana-3719 107 1 the the DET cana-3719 107 2 author author NOUN cana-3719 107 3 li li PROPN cana-3719 107 4 et et PROPN cana-3719 107 5 al al PROPN cana-3719 107 6 . . PUNCT cana-3719 108 1 [ [ X cana-3719 108 2 32 32 NUM cana-3719 108 3 ] ] PUNCT cana-3719 108 4 analyzed analyze VERB cana-3719 108 5 the the DET cana-3719 108 6 inability inability NOUN cana-3719 108 7 of of ADP cana-3719 108 8 the the DET cana-3719 108 9 traditional traditional ADJ cana-3719 108 10 gan gan NOUN cana-3719 108 11 in in ADP cana-3719 108 12 reconstructing reconstruct VERB cana-3719 108 13 the the DET cana-3719 108 14 texture texture NOUN cana-3719 108 15 , , PUNCT cana-3719 108 16 and and CCONJ cana-3719 108 17 structures structure NOUN cana-3719 108 18 of of ADP cana-3719 108 19 the the DET cana-3719 108 20 missed miss VERB cana-3719 108 21 region region NOUN cana-3719 108 22 simultaneously simultaneously ADV cana-3719 108 23 . . PUNCT cana-3719 109 1 this this DET cana-3719 109 2 approach approach NOUN cana-3719 109 3 introduces introduce VERB cana-3719 109 4 a a DET cana-3719 109 5 texture texture NOUN cana-3719 109 6 , , PUNCT cana-3719 109 7 and and CCONJ cana-3719 109 8 structureguided structureguide VERB cana-3719 109 9 model model NOUN cana-3719 109 10 , , PUNCT cana-3719 109 11 that that PRON cana-3719 109 12 uses use VERB cana-3719 109 13 the the DET cana-3719 109 14 subnetworks subnetwork NOUN cana-3719 109 15 such such ADJ cana-3719 109 16 as as ADP cana-3719 109 17 refinement refinement NOUN cana-3719 109 18 network network NOUN cana-3719 109 19 , , PUNCT cana-3719 109 20 texture texture NOUN cana-3719 109 21 , , PUNCT cana-3719 109 22 and and CCONJ cana-3719 109 23 structure structure NOUN cana-3719 109 24 reconstruction reconstruction NOUN cana-3719 109 25 process process NOUN cana-3719 109 26 . . PUNCT cana-3719 110 1 the the DET cana-3719 110 2 texture texture NOUN cana-3719 110 3 , , PUNCT cana-3719 110 4 and and CCONJ cana-3719 110 5 structure structure NOUN cana-3719 110 6 reconstruction reconstruction NOUN cana-3719 110 7 process process NOUN cana-3719 110 8 can can AUX cana-3719 110 9 restore restore VERB cana-3719 110 10 more more ADV cana-3719 110 11 fine fine ADJ cana-3719 110 12 texture texture NOUN cana-3719 110 13 , , PUNCT cana-3719 110 14 and and CCONJ cana-3719 110 15 coherent coherent ADJ cana-3719 110 16 structures structure NOUN cana-3719 110 17 than than ADP cana-3719 110 18 the the DET cana-3719 110 19 traditional traditional ADJ cana-3719 110 20 gan gan NOUN cana-3719 110 21 . . PUNCT cana-3719 111 1 finally finally ADV cana-3719 111 2 , , PUNCT cana-3719 111 3 the the DET cana-3719 111 4 refinement refinement NOUN cana-3719 111 5 network network NOUN cana-3719 111 6 combines combine VERB cana-3719 111 7 the the DET cana-3719 111 8 structure structure NOUN cana-3719 111 9 , , PUNCT cana-3719 111 10 and and CCONJ cana-3719 111 11 texture texture NOUN cana-3719 111 12 to to PART cana-3719 111 13 obtain obtain VERB cana-3719 111 14 the the DET cana-3719 111 15 inpainted inpainte VERB cana-3719 111 16 picture picture NOUN cana-3719 111 17 . . PUNCT cana-3719 112 1 to to PART cana-3719 112 2 reduce reduce VERB cana-3719 112 3 the the DET cana-3719 112 4 computational computational ADJ cana-3719 112 5 burden burden NOUN cana-3719 112 6 of of ADP cana-3719 112 7 the the DET cana-3719 112 8 gan gan PROPN cana-3719 112 9 structure structure NOUN cana-3719 112 10 , , PUNCT cana-3719 112 11 the the DET cana-3719 112 12 generator generator NOUN cana-3719 112 13 unit unit NOUN cana-3719 112 14 was be AUX cana-3719 112 15 included include VERB cana-3719 112 16 with with ADP cana-3719 112 17 a a DET cana-3719 112 18 multi multi ADJ cana-3719 112 19 - - ADJ cana-3719 112 20 task task ADJ cana-3719 112 21 learning learning NOUN cana-3719 112 22 process process NOUN cana-3719 112 23 [ [ X cana-3719 112 24 33 33 NUM cana-3719 112 25 ] ] PUNCT cana-3719 112 26 . . PUNCT cana-3719 113 1 this this DET cana-3719 113 2 approach approach NOUN cana-3719 113 3 was be AUX cana-3719 113 4 used use VERB cana-3719 113 5 in in ADP cana-3719 113 6 inpainting inpainte VERB cana-3719 113 7 the the DET cana-3719 113 8 biomedical biomedical ADJ cana-3719 113 9 images image NOUN cana-3719 113 10 in in ADP cana-3719 113 11 which which PRON cana-3719 113 12 the the DET cana-3719 113 13 encoder encoder NOUN cana-3719 113 14 - - PUNCT cana-3719 113 15 decoder decoder NOUN cana-3719 113 16 performs perform VERB cana-3719 113 17 three three NUM cana-3719 113 18 generator generator NOUN cana-3719 113 19 inpainting inpainte VERB cana-3719 113 20 target target NOUN cana-3719 113 21 fake fake ADJ cana-3719 113 22 image image NOUN cana-3719 113 23 real real ADJ cana-3719 113 24 image image NOUN cana-3719 113 25 discriminator discriminator NOUN cana-3719 113 26 fake fake ADJ cana-3719 113 27 or or CCONJ cana-3719 113 28 real real ADJ cana-3719 113 29 generator generator NOUN cana-3719 113 30 discriminator discriminator NOUN cana-3719 113 31 loss loss NOUN cana-3719 113 32 function function NOUN cana-3719 113 33 communications communication NOUN cana-3719 113 34 on on ADP cana-3719 113 35 applied apply VERB cana-3719 113 36 nonlinear nonlinear ADJ cana-3719 113 37 analysis analysis NOUN cana-3719 113 38 issn issn NOUN cana-3719 113 39 : : PUNCT cana-3719 113 40 1074 1074 NUM cana-3719 113 41 - - PUNCT cana-3719 113 42 133x 133x NUM cana-3719 113 43 vol vol NOUN cana-3719 113 44 32 32 NUM cana-3719 113 45 no no NOUN cana-3719 113 46 . . PUNCT cana-3719 114 1 8s 8s PROPN cana-3719 114 2 ( ( PUNCT cana-3719 114 3 2025 2025 NUM cana-3719 114 4 ) ) PUNCT cana-3719 114 5 581 581 NUM cana-3719 114 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 114 7 processes process NOUN cana-3719 114 8 namely namely ADV cana-3719 114 9 detecting detect VERB cana-3719 114 10 the the DET cana-3719 114 11 edges edge NOUN cana-3719 114 12 , , PUNCT cana-3719 114 13 generation generation NOUN cana-3719 114 14 of of ADP cana-3719 114 15 organ organ NOUN cana-3719 114 16 boundary boundary NOUN cana-3719 114 17 , , PUNCT cana-3719 114 18 and and CCONJ cana-3719 114 19 completion completion NOUN cana-3719 114 20 of of ADP cana-3719 114 21 resultant resultant NOUN cana-3719 114 22 images image NOUN cana-3719 114 23 . . PUNCT cana-3719 115 1 the the DET cana-3719 115 2 structure structure NOUN cana-3719 115 3 of of ADP cana-3719 115 4 the the DET cana-3719 115 5 gan gan PROPN cana-3719 115 6 uses use VERB cana-3719 115 7 regularization regularization NOUN cana-3719 115 8 functions function NOUN cana-3719 115 9 and and CCONJ cana-3719 115 10 an an DET cana-3719 115 11 optimum optimum ADJ cana-3719 115 12 number number NOUN cana-3719 115 13 of of ADP cana-3719 115 14 layers layer NOUN cana-3719 115 15 . . PUNCT cana-3719 116 1 a a DET cana-3719 116 2 dual dual ADJ cana-3719 116 3 - - PUNCT cana-3719 116 4 stage stage NOUN cana-3719 116 5 gan gan NOUN cana-3719 116 6 [ [ PUNCT cana-3719 116 7 34 34 NUM cana-3719 116 8 ] ] PUNCT cana-3719 116 9 was be AUX cana-3719 116 10 derived derive VERB cana-3719 116 11 where where SCONJ cana-3719 116 12 the the DET cana-3719 116 13 first first ADJ cana-3719 116 14 and and CCONJ cana-3719 116 15 second second ADJ cana-3719 116 16 stage stage NOUN cana-3719 116 17 uses use VERB cana-3719 116 18 structure structure NOUN cana-3719 116 19 - - PUNCT cana-3719 116 20 aware aware ADJ cana-3719 116 21 and and CCONJ cana-3719 116 22 texture texture ADJ cana-3719 116 23 - - PUNCT cana-3719 116 24 aware aware ADJ cana-3719 116 25 learning learning NOUN cana-3719 116 26 mechanisms mechanism NOUN cana-3719 116 27 . . PUNCT cana-3719 117 1 the the DET cana-3719 117 2 structure structure NOUN cana-3719 117 3 learning learning NOUN cana-3719 117 4 recovers recover VERB cana-3719 117 5 the the DET cana-3719 117 6 low low ADJ cana-3719 117 7 - - PUNCT cana-3719 117 8 frequency frequency NOUN cana-3719 117 9 component component NOUN cana-3719 117 10 , , PUNCT cana-3719 117 11 while while SCONJ cana-3719 117 12 the the DET cana-3719 117 13 texture texture NOUN cana-3719 117 14 - - PUNCT cana-3719 117 15 aware aware ADJ cana-3719 117 16 learning learning NOUN cana-3719 117 17 recovers recover VERB cana-3719 117 18 the the DET cana-3719 117 19 high high ADJ cana-3719 117 20 - - PUNCT cana-3719 117 21 frequency frequency NOUN cana-3719 117 22 component component NOUN cana-3719 117 23 . . PUNCT cana-3719 118 1 the the DET cana-3719 118 2 performance performance NOUN cana-3719 118 3 of of ADP cana-3719 118 4 gan gan PROPN cana-3719 118 5 was be AUX cana-3719 118 6 enhanced enhance VERB cana-3719 118 7 by by ADP cana-3719 118 8 models model NOUN cana-3719 118 9 namely namely ADV cana-3719 118 10 gan gan ADJ cana-3719 118 11 - - PUNCT cana-3719 118 12 gradient gradient NOUN cana-3719 118 13 penalty penalty NOUN cana-3719 118 14 [ [ X cana-3719 118 15 36 36 NUM cana-3719 118 16 ] ] PUNCT cana-3719 118 17 , , PUNCT cana-3719 118 18 least least ADJ cana-3719 118 19 squares square NOUN cana-3719 118 20 gan gan VERB cana-3719 119 1 [ [ PUNCT cana-3719 119 2 35 35 NUM cana-3719 119 3 ] ] PUNCT cana-3719 119 4 , , PUNCT cana-3719 119 5 and and CCONJ cana-3719 119 6 wasserstein wasserstein PROPN cana-3719 119 7 gan gan PROPN cana-3719 120 1 [ [ PUNCT cana-3719 120 2 37 37 NUM cana-3719 120 3 ] ] PUNCT cana-3719 120 4 architectures architecture NOUN cana-3719 120 5 . . PUNCT cana-3719 121 1 3.3 3.3 NUM cana-3719 121 2 transformer transformer NOUN cana-3719 121 3 based base VERB cana-3719 121 4 schmes schme NOUN cana-3719 121 5 fig fig NOUN cana-3719 121 6 . . PUNCT cana-3719 122 1 4 4 NUM cana-3719 122 2 : : PUNCT cana-3719 122 3 representation representation NOUN cana-3719 122 4 of of ADP cana-3719 122 5 transformer transformer NOUN cana-3719 122 6 model model NOUN cana-3719 122 7 in in ADP cana-3719 122 8 image image NOUN cana-3719 122 9 inpainting inpainte VERB cana-3719 122 10 fig fig NOUN cana-3719 122 11 . . PUNCT cana-3719 123 1 4 4 NUM cana-3719 123 2 shows show VERB cana-3719 123 3 the the DET cana-3719 123 4 structure structure NOUN cana-3719 123 5 of of ADP cana-3719 123 6 the the DET cana-3719 123 7 transformer transformer NOUN cana-3719 123 8 model model NOUN cana-3719 123 9 - - PUNCT cana-3719 123 10 based base VERB cana-3719 123 11 image image NOUN cana-3719 123 12 inpainting inpainte VERB cana-3719 123 13 . . PUNCT cana-3719 124 1 the the DET cana-3719 124 2 total total ADJ cana-3719 124 3 loss loss NOUN cana-3719 124 4 in in ADP cana-3719 124 5 transformer transformer NOUN cana-3719 124 6 - - PUNCT cana-3719 124 7 based base VERB cana-3719 124 8 schemes scheme NOUN cana-3719 124 9 is be AUX cana-3719 124 10 πœ‚π‘‘ πœ‚π‘‘ ADP cana-3719 124 11 = = PUNCT cana-3719 124 12 πœ†π‘π‘’π‘Ÿ πœ†π‘π‘’π‘Ÿ NOUN cana-3719 124 13 𝑁 𝑁 PROPN cana-3719 124 14 βˆ‘ βˆ‘ PUNCT cana-3719 124 15 ‖𝛿𝑛(𝐼𝑖 ‖𝛿𝑛(𝐼𝑖 PROPN cana-3719 124 16 ) ) PUNCT cana-3719 124 17 βˆ’ βˆ’ ADP cana-3719 124 18 𝛿𝑛(πΌπ‘Ÿ)β€–2 𝛿𝑛(πΌπ‘Ÿ)β€–2 NOUN cana-3719 124 19 2 2 NUM cana-3719 124 20 + + SYM cana-3719 124 21 1 1 NUM cana-3719 124 22 |𝜌| |𝜌| PROPN cana-3719 124 23 βˆ‘ βˆ‘ PUNCT cana-3719 124 24 β€–πΌπ‘Ÿ(𝑗 β€–πΌπ‘Ÿ(𝑗 ADV cana-3719 124 25 ) ) PUNCT cana-3719 124 26 βˆ’ βˆ’ PROPN cana-3719 125 1 𝐼𝑖(𝑗)β€–2 𝐼𝑖(𝑗)β€–2 ADV cana-3719 125 2 2 2 NUM cana-3719 125 3 π‘—βˆˆπœŒ π‘—βˆˆπœŒ NOUN cana-3719 125 4 𝑁 𝑁 NOUN cana-3719 125 5 𝑛=1 𝑛=1 NOUN cana-3719 125 6 ( ( PUNCT cana-3719 125 7 6 6 NUM cana-3719 125 8 ) ) PUNCT cana-3719 125 9 𝜌 𝜌 PART cana-3719 125 10 resembles resemble VERB cana-3719 125 11 the the DET cana-3719 125 12 indices index NOUN cana-3719 125 13 of of ADP cana-3719 125 14 masked mask VERB cana-3719 125 15 pixels pixel NOUN cana-3719 125 16 , , PUNCT cana-3719 125 17 𝛿𝑛 𝛿𝑛 PROPN cana-3719 125 18 ( ( PUNCT cana-3719 125 19 . . PUNCT cana-3719 125 20 ) ) PUNCT cana-3719 125 21 resembles resemble VERB cana-3719 125 22 the the DET cana-3719 125 23 features feature NOUN cana-3719 125 24 extracted extract VERB cana-3719 125 25 at at ADP cana-3719 125 26 the the DET cana-3719 125 27 nth nth NOUN cana-3719 125 28 layer layer NOUN cana-3719 125 29 . . PUNCT cana-3719 126 1 𝑁 𝑁 PROPN cana-3719 126 2 resembles resemble VERB cana-3719 126 3 the the DET cana-3719 126 4 number number NOUN cana-3719 126 5 of of ADP cana-3719 126 6 layers layer NOUN cana-3719 126 7 and and CCONJ cana-3719 126 8 πœ†π‘π‘’π‘Ÿ πœ†π‘π‘’π‘Ÿ NOUN cana-3719 126 9 resembles resemble VERB cana-3719 126 10 the the DET cana-3719 126 11 constant constant ADJ cana-3719 126 12 . . PUNCT cana-3719 127 1 the the DET cana-3719 127 2 transformer transformer ADJ cana-3719 127 3 model model NOUN cana-3719 127 4 [ [ X cana-3719 127 5 38 38 NUM cana-3719 127 6 ] ] PUNCT cana-3719 127 7 used use VERB cana-3719 127 8 by by ADP cana-3719 127 9 huang huang PROPN cana-3719 127 10 et et PROPN cana-3719 127 11 al al PROPN cana-3719 127 12 . . PROPN cana-3719 127 13 uses use VERB cana-3719 127 14 a a DET cana-3719 127 15 u u ADJ cana-3719 127 16 - - ADJ cana-3719 127 17 net net ADJ cana-3719 127 18 structured structured ADJ cana-3719 127 19 transformer transformer NOUN cana-3719 127 20 named name VERB cana-3719 127 21 self self NOUN cana-3719 127 22 - - PUNCT cana-3719 127 23 attention attention NOUN cana-3719 127 24 transformer transformer NOUN cana-3719 127 25 . . PUNCT cana-3719 128 1 this this DET cana-3719 128 2 transformer transformer NOUN cana-3719 128 3 reduces reduce VERB cana-3719 128 4 the the DET cana-3719 128 5 complexity complexity NOUN cana-3719 128 6 of of ADP cana-3719 128 7 the the DET cana-3719 128 8 multi multi ADJ cana-3719 128 9 - - ADJ cana-3719 128 10 headed headed ADJ cana-3719 128 11 attention attention NOUN cana-3719 128 12 structure structure NOUN cana-3719 128 13 . . PUNCT cana-3719 129 1 this this DET cana-3719 129 2 approach approach NOUN cana-3719 129 3 replaces replace VERB cana-3719 129 4 the the DET cana-3719 129 5 traditional traditional ADJ cana-3719 129 6 attention attention NOUN cana-3719 129 7 estimation estimation NOUN cana-3719 129 8 . . PUNCT cana-3719 130 1 the the DET cana-3719 130 2 irrelevant irrelevant ADJ cana-3719 130 3 features feature NOUN cana-3719 130 4 are be AUX cana-3719 130 5 excluded exclude VERB cana-3719 130 6 by by ADP cana-3719 130 7 the the DET cana-3719 130 8 sparse sparse ADJ cana-3719 130 9 attention attention NOUN cana-3719 130 10 map map NOUN cana-3719 130 11 that that PRON cana-3719 130 12 uses use VERB cana-3719 130 13 the the DET cana-3719 130 14 relu relu NOUN cana-3719 130 15 function function NOUN cana-3719 130 16 instead instead ADV cana-3719 130 17 of of ADP cana-3719 130 18 the the DET cana-3719 130 19 canonical canonical ADJ cana-3719 130 20 function function NOUN cana-3719 130 21 . . PUNCT cana-3719 131 1 transformer transformer NOUN cana-3719 131 2 - - PUNCT cana-3719 131 3 based base VERB cana-3719 131 4 schemes scheme NOUN cana-3719 131 5 such such ADJ cana-3719 131 6 as as ADP cana-3719 131 7 the the DET cana-3719 131 8 zits zit NOUN cana-3719 131 9 network network NOUN cana-3719 131 10 [ [ X cana-3719 131 11 39 39 NUM cana-3719 131 12 ] ] PUNCT cana-3719 131 13 , , PUNCT cana-3719 131 14 mae mae PROPN cana-3719 131 15 model model NOUN cana-3719 131 16 [ [ X cana-3719 131 17 40 40 NUM cana-3719 131 18 ] ] PUNCT cana-3719 131 19 , , PUNCT cana-3719 131 20 and and CCONJ cana-3719 131 21 transfill transfill ADV cana-3719 131 22 model model NOUN cana-3719 131 23 [ [ X cana-3719 131 24 41 41 NUM cana-3719 131 25 ] ] PUNCT cana-3719 131 26 also also ADV cana-3719 131 27 provide provide VERB cana-3719 131 28 reasonable reasonable ADJ cana-3719 131 29 performance performance NOUN cana-3719 131 30 . . PUNCT cana-3719 132 1 the the DET cana-3719 132 2 performance performance NOUN cana-3719 132 3 of of ADP cana-3719 132 4 the the DET cana-3719 132 5 transformer transformer NOUN cana-3719 132 6 reduces reduce VERB cana-3719 132 7 if if SCONJ cana-3719 132 8 the the DET cana-3719 132 9 missed miss VERB cana-3719 132 10 region region NOUN cana-3719 132 11 is be AUX cana-3719 132 12 large large ADJ cana-3719 132 13 . . PUNCT cana-3719 133 1 to to PART cana-3719 133 2 address address VERB cana-3719 133 3 this this DET cana-3719 133 4 challenge challenge NOUN cana-3719 133 5 , , PUNCT cana-3719 133 6 a a DET cana-3719 133 7 visual visual ADJ cana-3719 133 8 transformer transformer NOUN cana-3719 133 9 structure structure NOUN cana-3719 133 10 [ [ X cana-3719 133 11 42 42 NUM cana-3719 133 12 ] ] PUNCT cana-3719 133 13 is be AUX cana-3719 133 14 introduced introduce VERB cana-3719 133 15 that that PRON cana-3719 133 16 uses use VERB cana-3719 133 17 a a DET cana-3719 133 18 variable variable ADJ cana-3719 133 19 hyperparameter hyperparameter NOUN cana-3719 133 20 . . PUNCT cana-3719 134 1 this this DET cana-3719 134 2 approach approach NOUN cana-3719 134 3 initially initially ADV cana-3719 134 4 obtains obtain VERB cana-3719 134 5 multi multi ADJ cana-3719 134 6 - - ADJ cana-3719 134 7 scale scale ADJ cana-3719 134 8 patches patch NOUN cana-3719 134 9 by by ADP cana-3719 134 10 partitioning partition VERB cana-3719 134 11 the the DET cana-3719 134 12 descriptor descriptor NOUN cana-3719 134 13 maps map NOUN cana-3719 134 14 . . PUNCT cana-3719 135 1 the the DET cana-3719 135 2 computation computation NOUN cana-3719 135 3 burden burden NOUN cana-3719 135 4 of of ADP cana-3719 135 5 the the DET cana-3719 135 6 self self NOUN cana-3719 135 7 - - PUNCT cana-3719 135 8 attention attention NOUN cana-3719 135 9 process process NOUN cana-3719 135 10 is be AUX cana-3719 135 11 balanced balance VERB cana-3719 135 12 by by ADP cana-3719 135 13 assigning assign VERB cana-3719 135 14 the the DET cana-3719 135 15 feature feature NOUN cana-3719 135 16 map map NOUN cana-3719 135 17 to to ADP cana-3719 135 18 different different ADJ cana-3719 135 19 heads head NOUN cana-3719 135 20 . . PUNCT cana-3719 136 1 this this DET cana-3719 136 2 approach approach NOUN cana-3719 136 3 also also ADV cana-3719 136 4 minimizes minimize VERB cana-3719 136 5 the the DET cana-3719 136 6 number number NOUN cana-3719 136 7 of of ADP cana-3719 136 8 feature feature NOUN cana-3719 136 9 map map NOUN cana-3719 136 10 channels channel NOUN cana-3719 136 11 by by ADP cana-3719 136 12 a a DET cana-3719 136 13 depth depth NOUN cana-3719 136 14 - - PUNCT cana-3719 136 15 wise wise ADJ cana-3719 136 16 convolution convolution NOUN cana-3719 136 17 strategy strategy NOUN cana-3719 136 18 . . PUNCT cana-3719 137 1 a a DET cana-3719 137 2 dual dual ADJ cana-3719 137 3 discriminator discriminator NOUN cana-3719 137 4 approach approach NOUN cana-3719 137 5 [ [ X cana-3719 137 6 43 43 NUM cana-3719 137 7 ] ] PUNCT cana-3719 137 8 was be AUX cana-3719 137 9 used use VERB cana-3719 137 10 to to PART cana-3719 137 11 improve improve VERB cana-3719 137 12 the the DET cana-3719 137 13 local local ADJ cana-3719 137 14 and and CCONJ cana-3719 137 15 global global ADJ cana-3719 137 16 discrimination discrimination NOUN cana-3719 137 17 in in ADP cana-3719 137 18 gan gan PROPN cana-3719 137 19 . . PUNCT cana-3719 138 1 also also ADV cana-3719 138 2 , , PUNCT cana-3719 138 3 the the DET cana-3719 138 4 texture texture ADJ cana-3719 138 5 - - PUNCT cana-3719 138 6 aware aware ADJ cana-3719 138 7 module module NOUN cana-3719 138 8 in in ADP cana-3719 138 9 gan gan ADJ cana-3719 138 10 architecture architecture NOUN cana-3719 138 11 [ [ X cana-3719 138 12 44 44 NUM cana-3719 138 13 ] ] PUNCT cana-3719 138 14 reduces reduce VERB cana-3719 138 15 the the DET cana-3719 138 16 complexity complexity NOUN cana-3719 138 17 of of ADP cana-3719 138 18 the the DET cana-3719 138 19 traditional traditional ADJ cana-3719 138 20 gan gan ADJ cana-3719 138 21 structure structure NOUN cana-3719 138 22 . . PUNCT cana-3719 139 1 cnn cnn PROPN cana-3719 139 2 p p NOUN cana-3719 139 3 at at ADP cana-3719 139 4 ch ch PROPN cana-3719 139 5 e e PROPN cana-3719 139 6 m m PROPN cana-3719 139 7 b b NOUN cana-3719 139 8 ed ed NOUN cana-3719 139 9 d d PROPN cana-3719 139 10 in in ADP cana-3719 139 11 g g PROPN cana-3719 139 12 p p X cana-3719 139 13 o o X cana-3719 139 14 si si INTJ cana-3719 139 15 ti ti X cana-3719 139 16 o o PROPN cana-3719 139 17 n n PROPN cana-3719 139 18 e e NOUN cana-3719 139 19 m m PROPN cana-3719 139 20 b b NOUN cana-3719 140 1 ed ed NOUN cana-3719 141 1 d d PROPN cana-3719 141 2 in in ADP cana-3719 141 3 g g PROPN cana-3719 141 4 0 0 NUM cana-3719 141 5 1 1 NUM cana-3719 141 6 t t NOUN cana-3719 141 7 ra ra PROPN cana-3719 141 8 n n ADV cana-3719 141 9 sf sf NOUN cana-3719 141 10 o o PROPN cana-3719 141 11 rm rm NOUN cana-3719 141 12 er er INTJ cana-3719 141 13 e e PROPN cana-3719 141 14 n n X cana-3719 141 15 co co NOUN cana-3719 141 16 d d PROPN cana-3719 141 17 er er INTJ cana-3719 141 18 0 0 NUM cana-3719 141 19 1 1 NUM cana-3719 141 20 u u NOUN cana-3719 141 21 n n PRON cana-3719 141 22 fl fl NOUN cana-3719 141 23 at at ADP cana-3719 141 24 te te PROPN cana-3719 141 25 n n PRON cana-3719 141 26 p p NOUN cana-3719 141 27 o o X cana-3719 141 28 si si INTJ cana-3719 141 29 ti ti X cana-3719 141 30 o o PROPN cana-3719 141 31 n n PROPN cana-3719 141 32 e e NOUN cana-3719 141 33 m m PROPN cana-3719 141 34 b b NOUN cana-3719 141 35 ed ed NOUN cana-3719 142 1 d d PROPN cana-3719 142 2 in in ADP cana-3719 142 3 g g PROPN cana-3719 142 4 encoder encoder NOUN cana-3719 143 1 d d ADP cana-3719 143 2 ec ec PROPN cana-3719 144 1 o o NOUN cana-3719 144 2 d d X cana-3719 144 3 er er INTJ cana-3719 144 4 communications communication NOUN cana-3719 144 5 on on ADP cana-3719 144 6 applied apply VERB cana-3719 144 7 nonlinear nonlinear ADJ cana-3719 144 8 analysis analysis NOUN cana-3719 144 9 issn issn NOUN cana-3719 144 10 : : PUNCT cana-3719 144 11 1074 1074 NUM cana-3719 144 12 - - PUNCT cana-3719 144 13 133x 133x NUM cana-3719 144 14 vol vol NOUN cana-3719 144 15 32 32 NUM cana-3719 144 16 no no NOUN cana-3719 144 17 . . PUNCT cana-3719 145 1 8s 8s PROPN cana-3719 145 2 ( ( PUNCT cana-3719 145 3 2025 2025 NUM cana-3719 145 4 ) ) PUNCT cana-3719 145 5 582 582 NUM cana-3719 145 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 145 7 3.4 3.4 NUM cana-3719 145 8 attention attention NOUN cana-3719 145 9 - - PUNCT cana-3719 145 10 based base VERB cana-3719 145 11 schemes scheme NOUN cana-3719 145 12 fig fig NOUN cana-3719 145 13 . . PUNCT cana-3719 146 1 5 5 NUM cana-3719 146 2 : : PUNCT cana-3719 146 3 representation representation NOUN cana-3719 146 4 of of ADP cana-3719 146 5 attention attention NOUN cana-3719 146 6 - - PUNCT cana-3719 146 7 based base VERB cana-3719 146 8 image image NOUN cana-3719 146 9 painting paint VERB cana-3719 146 10 the the DET cana-3719 146 11 attention attention NOUN cana-3719 146 12 - - PUNCT cana-3719 146 13 based base VERB cana-3719 146 14 approach approach NOUN cana-3719 146 15 provides provide VERB cana-3719 146 16 a a DET cana-3719 146 17 high high ADJ cana-3719 146 18 preference preference NOUN cana-3719 146 19 for for ADP cana-3719 146 20 certain certain ADJ cana-3719 146 21 regions region NOUN cana-3719 146 22 while while SCONJ cana-3719 146 23 reconstructing reconstruct VERB cana-3719 146 24 the the DET cana-3719 146 25 missed miss VERB cana-3719 146 26 region region NOUN cana-3719 146 27 . . PUNCT cana-3719 147 1 fig fig NOUN cana-3719 147 2 . . PUNCT cana-3719 148 1 5 5 NUM cana-3719 148 2 illustrates illustrate VERB cana-3719 148 3 the the DET cana-3719 148 4 representation representation NOUN cana-3719 148 5 of of ADP cana-3719 148 6 attention attention NOUN cana-3719 148 7 - - PUNCT cana-3719 148 8 based base VERB cana-3719 148 9 image image NOUN cana-3719 148 10 painting painting NOUN cana-3719 148 11 [ [ X cana-3719 148 12 24 24 NUM cana-3719 148 13 ] ] PUNCT cana-3719 148 14 which which PRON cana-3719 148 15 has have VERB cana-3719 148 16 the the DET cana-3719 148 17 single single ADJ cana-3719 148 18 encoder encoder NOUN cana-3719 148 19 section section NOUN cana-3719 148 20 with with ADP cana-3719 148 21 convolutional convolutional ADJ cana-3719 148 22 filters filter NOUN cana-3719 148 23 and and CCONJ cana-3719 148 24 resblocks resblock NOUN cana-3719 148 25 represented represent VERB cana-3719 148 26 as as ADP cana-3719 148 27 ( ( PUNCT cana-3719 148 28 𝑅𝑒𝑛 𝑅𝑒𝑛 NOUN cana-3719 148 29 ) ) PUNCT cana-3719 148 30 , , PUNCT cana-3719 148 31 while while SCONJ cana-3719 148 32 the the DET cana-3719 148 33 decoder decoder NOUN cana-3719 148 34 has have VERB cana-3719 148 35 three three NUM cana-3719 148 36 sections section NOUN cana-3719 148 37 of of ADP cana-3719 148 38 convolutional convolutional ADJ cana-3719 148 39 filters filter NOUN cana-3719 148 40 and and CCONJ cana-3719 148 41 attention attention NOUN cana-3719 148 42 modules module NOUN cana-3719 148 43 represented represent VERB cana-3719 148 44 as as ADP cana-3719 148 45 ( ( PUNCT cana-3719 148 46 𝑅𝑑𝑒 𝑅𝑑𝑒 NOUN cana-3719 148 47 1 1 NUM cana-3719 148 48 , , PUNCT cana-3719 148 49 𝑅𝑑𝑒 𝑅𝑑𝑒 NOUN cana-3719 148 50 2 2 NUM cana-3719 148 51 , , PUNCT cana-3719 148 52 𝑅𝑑𝑒 𝑅𝑑𝑒 PROPN cana-3719 148 53 3 3 NUM cana-3719 148 54 ) ) PUNCT cana-3719 148 55 . . PUNCT cana-3719 149 1 two two NUM cana-3719 149 2 attention attention NOUN cana-3719 149 3 modules module NOUN cana-3719 149 4 namely namely ADV cana-3719 149 5 multiscale multiscale VERB cana-3719 149 6 attention attention NOUN cana-3719 149 7 - - PUNCT cana-3719 149 8 based base VERB cana-3719 149 9 feature feature NOUN cana-3719 149 10 extraction extraction NOUN cana-3719 149 11 ( ( PUNCT cana-3719 149 12 mafe mafe ADJ cana-3719 149 13 ) ) PUNCT cana-3719 149 14 and and CCONJ cana-3719 149 15 gradient gradient ADJ cana-3719 149 16 attention attention NOUN cana-3719 149 17 guidance guidance NOUN cana-3719 149 18 ( ( PUNCT cana-3719 149 19 gag gag NOUN cana-3719 149 20 ) ) PUNCT cana-3719 149 21 are be AUX cana-3719 149 22 utilized utilize VERB cana-3719 149 23 in in ADP cana-3719 149 24 the the DET cana-3719 149 25 three three NUM cana-3719 149 26 sections section NOUN cana-3719 149 27 of of ADP cana-3719 149 28 the the DET cana-3719 149 29 decoder decoder NOUN cana-3719 149 30 . . PUNCT cana-3719 150 1 four four NUM cana-3719 150 2 different different ADJ cana-3719 150 3 loss loss NOUN cana-3719 150 4 functions function NOUN cana-3719 150 5 namely namely ADV cana-3719 150 6 style style NOUN cana-3719 150 7 loss loss NOUN cana-3719 150 8 ( ( PUNCT cana-3719 150 9 𝐿𝑠𝑑𝑦𝑙𝑒 𝐿𝑠𝑑𝑦𝑙𝑒 PROPN cana-3719 150 10 ) ) PUNCT cana-3719 150 11 , , PUNCT cana-3719 150 12 perception perception NOUN cana-3719 150 13 loss loss NOUN cana-3719 150 14 ( ( PUNCT cana-3719 150 15 𝐿 𝐿 NOUN cana-3719 150 16 π‘π‘’π‘Ÿπ‘ π‘π‘’π‘Ÿπ‘ NOUN cana-3719 150 17 ) ) PUNCT cana-3719 150 18 , , PUNCT cana-3719 150 19 reconstruction reconstruction NOUN cana-3719 150 20 loss loss NOUN cana-3719 150 21 ( ( PUNCT cana-3719 150 22 𝐿 𝐿 PROPN cana-3719 150 23 π‘Ÿπ‘’π‘ π‘Ÿπ‘’π‘ NOUN cana-3719 150 24 ) ) PUNCT cana-3719 150 25 and and CCONJ cana-3719 150 26 gradient gradient ADJ cana-3719 150 27 loss loss NOUN cana-3719 150 28 ( ( PUNCT cana-3719 150 29 𝐿𝐺 𝐿𝐺 PROPN cana-3719 150 30 ) ) PUNCT cana-3719 150 31 are be AUX cana-3719 150 32 utilized utilize VERB cana-3719 150 33 to to PART cana-3719 150 34 update update VERB cana-3719 150 35 the the DET cana-3719 150 36 multiscale multiscale ADJ cana-3719 150 37 gradient gradient NOUN cana-3719 150 38 loss loss NOUN cana-3719 150 39 ( ( PUNCT cana-3719 150 40 π›₯𝐺 π›₯𝐺 PROPN cana-3719 150 41 ) ) PUNCT cana-3719 150 42 of of ADP cana-3719 150 43 the the DET cana-3719 150 44 decoder decoder NOUN cana-3719 150 45 . . PUNCT cana-3719 151 1 different different ADJ cana-3719 151 2 attention attention NOUN cana-3719 151 3 approaches approach VERB cana-3719 151 4 namely namely ADV cana-3719 151 5 coherent coherent ADJ cana-3719 151 6 semantic semantic ADJ cana-3719 151 7 attention attention NOUN cana-3719 152 1 [ [ X cana-3719 152 2 46 46 NUM cana-3719 152 3 ] ] PUNCT cana-3719 152 4 , , PUNCT cana-3719 152 5 contextual contextual ADJ cana-3719 152 6 attention attention NOUN cana-3719 153 1 [ [ X cana-3719 153 2 45 45 NUM cana-3719 153 3 ] ] PUNCT cana-3719 153 4 , , PUNCT cana-3719 153 5 and and CCONJ cana-3719 153 6 multistage multistage VERB cana-3719 153 7 attention attention NOUN cana-3719 153 8 [ [ X cana-3719 153 9 47 47 NUM cana-3719 153 10 ] ] PUNCT cana-3719 153 11 provide provide VERB cana-3719 153 12 fine fine ADJ cana-3719 153 13 texture texture ADJ cana-3719 153 14 information information NOUN cana-3719 153 15 during during ADP cana-3719 153 16 the the DET cana-3719 153 17 recovery recovery NOUN cana-3719 153 18 of of ADP cana-3719 153 19 missed miss VERB cana-3719 153 20 regions region NOUN cana-3719 153 21 . . PUNCT cana-3719 154 1 3.5 3.5 NUM cana-3719 154 2 u u NOUN cana-3719 154 3 - - NOUN cana-3719 154 4 net net ADJ cana-3719 154 5 based base VERB cana-3719 154 6 schemes scheme NOUN cana-3719 154 7 the the DET cana-3719 154 8 structure structure NOUN cana-3719 154 9 of of ADP cana-3719 154 10 the the DET cana-3719 154 11 u u NOUN cana-3719 154 12 - - NOUN cana-3719 154 13 net net ADJ cana-3719 154 14 and and CCONJ cana-3719 154 15 encoder encoder NOUN cana-3719 154 16 - - PUNCT cana-3719 154 17 decoder decoder NOUN cana-3719 154 18 structures structure NOUN cana-3719 154 19 are be AUX cana-3719 154 20 represented represent VERB cana-3719 154 21 in in ADP cana-3719 154 22 fig fig NOUN cana-3719 154 23 . . PUNCT cana-3719 155 1 6 6 NUM cana-3719 155 2 . . X cana-3719 155 3 deep deep ADJ cana-3719 155 4 networks network NOUN cana-3719 155 5 and and CCONJ cana-3719 155 6 attention attention NOUN cana-3719 155 7 processes process NOUN cana-3719 155 8 are be AUX cana-3719 155 9 combined combine VERB cana-3719 155 10 [ [ X cana-3719 155 11 48 48 NUM cana-3719 155 12 ] ] PUNCT cana-3719 155 13 to to PART cana-3719 155 14 improve improve VERB cana-3719 155 15 the the DET cana-3719 155 16 texture texture ADJ cana-3719 155 17 distortion distortion NOUN cana-3719 155 18 during during ADP cana-3719 155 19 the the DET cana-3719 155 20 reconstruction reconstruction NOUN cana-3719 155 21 process process NOUN cana-3719 155 22 . . PUNCT cana-3719 156 1 the the DET cana-3719 156 2 generator generator NOUN cana-3719 156 3 was be AUX cana-3719 156 4 constructed construct VERB cana-3719 156 5 using use VERB cana-3719 156 6 the the DET cana-3719 156 7 res re NOUN cana-3719 156 8 - - PUNCT cana-3719 156 9 u u NOUN cana-3719 156 10 - - ADJ cana-3719 156 11 net net ADJ cana-3719 156 12 structure structure NOUN cana-3719 156 13 . . PUNCT cana-3719 157 1 the the DET cana-3719 157 2 residual residual ADJ cana-3719 157 3 structure structure NOUN cana-3719 157 4 was be AUX cana-3719 157 5 used use VERB cana-3719 157 6 in in ADP cana-3719 157 7 the the DET cana-3719 157 8 encoder encoder NOUN cana-3719 157 9 of of ADP cana-3719 157 10 the the DET cana-3719 157 11 u u NOUN cana-3719 157 12 - - ADJ cana-3719 157 13 net net ADJ cana-3719 157 14 - - PUNCT cana-3719 157 15 based base VERB cana-3719 157 16 backbone backbone NOUN cana-3719 157 17 . . PUNCT cana-3719 158 1 the the DET cana-3719 158 2 skip skip ADJ cana-3719 158 3 connection connection NOUN cana-3719 158 4 uses use VERB cana-3719 158 5 the the DET cana-3719 158 6 attention attention NOUN cana-3719 158 7 module module NOUN cana-3719 158 8 which which PRON cana-3719 158 9 improves improve VERB cana-3719 158 10 the the DET cana-3719 158 11 effectiveness effectiveness NOUN cana-3719 158 12 of of ADP cana-3719 158 13 the the DET cana-3719 158 14 inpainting inpainting ADJ cana-3719 158 15 process process NOUN cana-3719 158 16 . . PUNCT cana-3719 159 1 fig fig NOUN cana-3719 159 2 . . PUNCT cana-3719 160 1 6 6 NUM cana-3719 160 2 : : PUNCT cana-3719 160 3 difference difference NOUN cana-3719 160 4 between between ADP cana-3719 160 5 u u NOUN cana-3719 160 6 - - ADJ cana-3719 160 7 net net ADJ cana-3719 160 8 architecture architecture NOUN cana-3719 160 9 and and CCONJ cana-3719 160 10 encoder encoder NOUN cana-3719 160 11 - - PUNCT cana-3719 160 12 decoder decoder NOUN cana-3719 160 13 structures structure NOUN cana-3719 160 14 the the DET cana-3719 160 15 loss loss NOUN cana-3719 160 16 function function NOUN cana-3719 160 17 in in ADP cana-3719 160 18 u u ADJ cana-3719 160 19 - - ADJ cana-3719 160 20 net net ADJ cana-3719 160 21 - - PUNCT cana-3719 160 22 based base VERB cana-3719 160 23 image image NOUN cana-3719 160 24 inpainting inpainte VERB cana-3719 160 25 can can AUX cana-3719 160 26 be be AUX cana-3719 160 27 represented represent VERB cana-3719 160 28 as as ADP cana-3719 160 29 , , PUNCT cana-3719 160 30 ƞ(πΌπ‘Ÿ ƞ(πΌπ‘Ÿ X cana-3719 160 31 , , PUNCT cana-3719 160 32 𝐼𝑖 𝐼𝑖 PROPN cana-3719 160 33 ) ) PUNCT cana-3719 160 34 = = SYM cana-3719 160 35 πœ†π‘‘Ζž πœ†π‘‘Ζž PRON cana-3719 160 36 𝑑 𝑑 X cana-3719 160 37 ( ( PUNCT cana-3719 160 38 πΌπ‘Ÿ πΌπ‘Ÿ PROPN cana-3719 160 39 , , PUNCT cana-3719 160 40 𝐼𝑖 𝐼𝑖 PROPN cana-3719 160 41 ) ) PUNCT cana-3719 160 42 + + NUM cana-3719 160 43 πœ†π‘πœ‚π‘(πΌπ‘Ÿ πœ†π‘πœ‚π‘(πΌπ‘Ÿ PROPN cana-3719 160 44 , , PUNCT cana-3719 160 45 𝐼𝑖 𝐼𝑖 PROPN cana-3719 160 46 ) ) PUNCT cana-3719 160 47 ( ( PUNCT cana-3719 160 48 7 7 X cana-3719 160 49 ) ) PUNCT cana-3719 160 50 here here ADV cana-3719 160 51 , , PUNCT cana-3719 160 52 πœ‚π‘ πœ‚π‘ NOUN cana-3719 160 53 resembles resemble VERB cana-3719 160 54 the the DET cana-3719 160 55 perceptual perceptual ADJ cana-3719 160 56 loss loss NOUN cana-3719 160 57 represented represent VERB cana-3719 160 58 as as ADP cana-3719 160 59 , , PUNCT cana-3719 160 60 communications communication NOUN cana-3719 160 61 on on ADP cana-3719 160 62 applied apply VERB cana-3719 160 63 nonlinear nonlinear ADJ cana-3719 160 64 analysis analysis NOUN cana-3719 160 65 issn issn NOUN cana-3719 160 66 : : PUNCT cana-3719 160 67 1074 1074 NUM cana-3719 160 68 - - PUNCT cana-3719 160 69 133x 133x NUM cana-3719 160 70 vol vol NOUN cana-3719 160 71 32 32 NUM cana-3719 160 72 no no NOUN cana-3719 160 73 . . PUNCT cana-3719 161 1 8s 8s PROPN cana-3719 161 2 ( ( PUNCT cana-3719 161 3 2025 2025 NUM cana-3719 161 4 ) ) PUNCT cana-3719 161 5 583 583 NUM cana-3719 161 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 161 7 ƞ ƞ SYM cana-3719 161 8 𝑝 𝑝 PROPN cana-3719 161 9 ( ( PUNCT cana-3719 161 10 πΌπ‘Ÿ πΌπ‘Ÿ PROPN cana-3719 161 11 , , PUNCT cana-3719 161 12 𝐼𝑖 𝐼𝑖 PROPN cana-3719 161 13 ) ) PUNCT cana-3719 161 14 = = SYM cana-3719 161 15 ‖𝛿(πΌπ‘Ÿ ‖𝛿(πΌπ‘Ÿ NOUN cana-3719 161 16 ) ) PUNCT cana-3719 161 17 βˆ’ βˆ’ NOUN cana-3719 161 18 𝛿(𝐼𝑖)β€–2 𝛿(𝐼𝑖)β€–2 NOUN cana-3719 161 19 2 2 NUM cana-3719 161 20 ( ( PUNCT cana-3719 161 21 8) 8) NUM cana-3719 161 22 here here ADV cana-3719 161 23 , , PUNCT cana-3719 161 24 𝛿(βˆ™ 𝛿(βˆ™ NOUN cana-3719 161 25 ) ) PUNCT cana-3719 161 26 represents represent VERB cana-3719 161 27 the the DET cana-3719 161 28 descriptor descriptor NOUN cana-3719 161 29 extraction extraction NOUN cana-3719 161 30 function function NOUN cana-3719 161 31 . . PUNCT cana-3719 162 1 πΌπ‘Ÿ πΌπ‘Ÿ ADP cana-3719 162 2 and and CCONJ cana-3719 162 3 𝐼𝑖 𝐼𝑖 PROPN cana-3719 162 4 resembles resemble VERB cana-3719 162 5 the the DET cana-3719 162 6 reconstructed reconstructed ADJ cana-3719 162 7 image image NOUN cana-3719 162 8 and and CCONJ cana-3719 162 9 input input NOUN cana-3719 162 10 image image NOUN cana-3719 162 11 respectively respectively ADV cana-3719 162 12 . . PUNCT cana-3719 163 1 in in ADP cana-3719 163 2 equation equation NOUN cana-3719 163 3 ( ( PUNCT cana-3719 163 4 7 7 NUM cana-3719 163 5 ) ) PUNCT cana-3719 163 6 , , PUNCT cana-3719 163 7 the the DET cana-3719 163 8 total total ADJ cana-3719 163 9 variation variation NOUN cana-3719 163 10 loss loss NOUN cana-3719 163 11 function function NOUN cana-3719 163 12 πœ‚π‘‘ πœ‚π‘‘ NOUN cana-3719 163 13 can can AUX cana-3719 163 14 be be AUX cana-3719 163 15 denoted denote VERB cana-3719 163 16 by by ADP cana-3719 163 17 , , PUNCT cana-3719 163 18 πœ‚π‘‘(πΌπ‘Ÿ πœ‚π‘‘(πΌπ‘Ÿ NOUN cana-3719 163 19 ) ) PUNCT cana-3719 163 20 = = SYM cana-3719 163 21 βˆ‘ βˆ‘ PUNCT cana-3719 163 22 { { PUNCT cana-3719 163 23 ( ( PUNCT cana-3719 163 24 |πΌπ‘Ÿ(𝑒 |πΌπ‘Ÿ(𝑒 NOUN cana-3719 163 25 + + CCONJ cana-3719 163 26 1 1 NUM cana-3719 163 27 , , PUNCT cana-3719 163 28 𝑣 𝑣 NOUN cana-3719 163 29 ) ) PUNCT cana-3719 163 30 βˆ’ βˆ’ PROPN cana-3719 163 31 πΌπ‘Ÿ(𝑒 πΌπ‘Ÿ(𝑒 NOUN cana-3719 163 32 , , PUNCT cana-3719 163 33 𝑣)| 𝑣)| NOUN cana-3719 163 34 ) ) PUNCT cana-3719 163 35 + + CCONJ cana-3719 163 36 ( ( PUNCT cana-3719 163 37 |πΌπ‘Ÿ(𝑒 |πΌπ‘Ÿ(𝑒 NOUN cana-3719 164 1 + + CCONJ cana-3719 164 2 1 1 NUM cana-3719 164 3 , , PUNCT cana-3719 164 4 𝑣 𝑣 NOUN cana-3719 164 5 ) ) PUNCT cana-3719 164 6 βˆ’ βˆ’ PROPN cana-3719 165 1 πΌπ‘Ÿ(𝑒 πΌπ‘Ÿ(𝑒 PROPN cana-3719 165 2 , , PUNCT cana-3719 165 3 𝑣)|)}𝑒,𝑣 𝑣)|)}𝑒,𝑣 PROPN cana-3719 165 4 ( ( PUNCT cana-3719 165 5 9 9 NUM cana-3719 165 6 ) ) PUNCT cana-3719 165 7 πœ†π‘ πœ†π‘ ADP cana-3719 165 8 , , PUNCT cana-3719 165 9 and and CCONJ cana-3719 165 10 πœ†π‘™ πœ†π‘™ NOUN cana-3719 165 11 are be AUX cana-3719 165 12 hyperparameters hyperparameter NOUN cana-3719 165 13 to to PART cana-3719 165 14 control control VERB cana-3719 165 15 the the DET cana-3719 165 16 perceptual perceptual NOUN cana-3719 165 17 , , PUNCT cana-3719 165 18 and and CCONJ cana-3719 165 19 total total ADJ cana-3719 165 20 variation variation NOUN cana-3719 165 21 loss loss NOUN cana-3719 165 22 respectively respectively ADV cana-3719 165 23 . . PUNCT cana-3719 166 1 different different ADJ cana-3719 166 2 networks network NOUN cana-3719 166 3 are be AUX cana-3719 166 4 derived derive VERB cana-3719 166 5 from from ADP cana-3719 166 6 the the DET cana-3719 166 7 u u ADJ cana-3719 166 8 - - ADJ cana-3719 166 9 net net ADJ cana-3719 166 10 structure structure NOUN cana-3719 166 11 that that PRON cana-3719 166 12 involves involve VERB cana-3719 166 13 pyramid pyramid NOUN cana-3719 166 14 - - PUNCT cana-3719 166 15 network network NOUN cana-3719 166 16 [ [ X cana-3719 166 17 49 49 NUM cana-3719 166 18 ] ] PUNCT cana-3719 166 19 , , PUNCT cana-3719 166 20 partial partial ADJ cana-3719 166 21 convolution convolution NOUN cana-3719 166 22 [ [ X cana-3719 166 23 50 50 NUM cana-3719 166 24 ] ] PUNCT cana-3719 166 25 , , PUNCT cana-3719 166 26 deep deep ADJ cana-3719 166 27 fusion fusion NOUN cana-3719 166 28 network network NOUN cana-3719 166 29 [ [ X cana-3719 166 30 51 51 NUM cana-3719 166 31 ] ] PUNCT cana-3719 166 32 , , PUNCT cana-3719 166 33 and and CCONJ cana-3719 166 34 shift shift NOUN cana-3719 166 35 - - PUNCT cana-3719 166 36 net net NOUN cana-3719 166 37 [ [ X cana-3719 166 38 52 52 NUM cana-3719 166 39 ] ] PUNCT cana-3719 166 40 . . PUNCT cana-3719 167 1 the the DET cana-3719 167 2 block block NOUN cana-3719 167 3 attention attention NOUN cana-3719 167 4 process process NOUN cana-3719 167 5 was be AUX cana-3719 167 6 included include VERB cana-3719 167 7 with with ADP cana-3719 167 8 the the DET cana-3719 167 9 resnet resnet ADJ cana-3719 167 10 architecture architecture NOUN cana-3719 167 11 to to PART cana-3719 167 12 derive derive VERB cana-3719 167 13 the the DET cana-3719 167 14 improved improved ADJ cana-3719 167 15 inpainting inpainte VERB cana-3719 167 16 architecture architecture NOUN cana-3719 167 17 [ [ X cana-3719 167 18 53 53 NUM cana-3719 167 19 ] ] PUNCT cana-3719 167 20 that that PRON cana-3719 167 21 shows show VERB cana-3719 167 22 a a DET cana-3719 167 23 lesser less ADJ cana-3719 167 24 variation variation NOUN cana-3719 167 25 between between ADP cana-3719 167 26 the the DET cana-3719 167 27 boundaries boundary NOUN cana-3719 167 28 of of ADP cana-3719 167 29 the the DET cana-3719 167 30 mask mask NOUN cana-3719 167 31 region region NOUN cana-3719 167 32 and and CCONJ cana-3719 167 33 the the DET cana-3719 167 34 un un PROPN cana-3719 167 35 - - PROPN cana-3719 167 36 mask mask NOUN cana-3719 167 37 region region NOUN cana-3719 167 38 . . PUNCT cana-3719 168 1 the the DET cana-3719 168 2 selfattention selfattention NOUN cana-3719 168 3 - - PUNCT cana-3719 168 4 based base VERB cana-3719 168 5 symmetric symmetric ADJ cana-3719 168 6 connected connect VERB cana-3719 168 7 u u NOUN cana-3719 168 8 - - NOUN cana-3719 168 9 net net ADJ cana-3719 168 10 [ [ X cana-3719 168 11 54 54 NUM cana-3719 168 12 ] ] PUNCT cana-3719 168 13 estimates estimate VERB cana-3719 168 14 fine fine ADJ cana-3719 168 15 texture texture NOUN cana-3719 168 16 on on ADP cana-3719 168 17 the the DET cana-3719 168 18 missed miss VERB cana-3719 168 19 regions region NOUN cana-3719 168 20 . . PUNCT cana-3719 169 1 3.6 3.6 NUM cana-3719 169 2 texture texture NOUN cana-3719 169 3 and and CCONJ cana-3719 169 4 structure structure NOUN cana-3719 169 5 based base VERB cana-3719 169 6 schemes scheme NOUN cana-3719 169 7 in in ADP cana-3719 169 8 this this DET cana-3719 169 9 scheme scheme NOUN cana-3719 169 10 , , PUNCT cana-3719 169 11 the the DET cana-3719 169 12 texture texture NOUN cana-3719 169 13 of of ADP cana-3719 169 14 the the DET cana-3719 169 15 unmasked unmasked ADJ cana-3719 169 16 area area NOUN cana-3719 169 17 is be AUX cana-3719 169 18 utilized utilize VERB cana-3719 169 19 to to PART cana-3719 169 20 fill fill VERB cana-3719 169 21 the the DET cana-3719 169 22 masked masked ADJ cana-3719 169 23 area area NOUN cana-3719 169 24 . . PUNCT cana-3719 170 1 the the DET cana-3719 170 2 texture texture ADJ cana-3719 170 3 pattern pattern NOUN cana-3719 170 4 of of ADP cana-3719 170 5 the the DET cana-3719 170 6 masked mask VERB cana-3719 170 7 area area NOUN cana-3719 170 8 is be AUX cana-3719 170 9 synthesized synthesize VERB cana-3719 170 10 based base VERB cana-3719 170 11 on on ADP cana-3719 170 12 the the DET cana-3719 170 13 structure structure NOUN cana-3719 170 14 , , PUNCT cana-3719 170 15 and and CCONJ cana-3719 170 16 texture texture NOUN cana-3719 170 17 of of ADP cana-3719 170 18 neighboring neighbor VERB cana-3719 170 19 unmasked unmasked ADJ cana-3719 170 20 pixels pixel NOUN cana-3719 170 21 . . PUNCT cana-3719 171 1 wang wang PROPN cana-3719 171 2 et et PROPN cana-3719 171 3 al al PROPN cana-3719 171 4 . . PUNCT cana-3719 172 1 [ [ X cana-3719 172 2 55 55 NUM cana-3719 172 3 ] ] PUNCT cana-3719 172 4 used use VERB cana-3719 172 5 a a DET cana-3719 172 6 small small ADJ cana-3719 172 7 receptive receptive ADJ cana-3719 172 8 field field NOUN cana-3719 172 9 - - PUNCT cana-3719 172 10 based base VERB cana-3719 172 11 shallow shallow NOUN cana-3719 172 12 , , PUNCT cana-3719 172 13 and and CCONJ cana-3719 172 14 deep deep ADJ cana-3719 172 15 structure structure NOUN cana-3719 172 16 to to PART cana-3719 172 17 minimize minimize VERB cana-3719 172 18 the the DET cana-3719 172 19 formation formation NOUN cana-3719 172 20 of of ADP cana-3719 172 21 artifacts artifact NOUN cana-3719 172 22 during during ADP cana-3719 172 23 reconstruction reconstruction NOUN cana-3719 172 24 and and CCONJ cana-3719 172 25 to to PART cana-3719 172 26 perform perform VERB cana-3719 172 27 local local ADJ cana-3719 172 28 refinement refinement NOUN cana-3719 172 29 . . PUNCT cana-3719 173 1 the the DET cana-3719 173 2 global global PROPN cana-3719 173 3 refinement refinement NOUN cana-3719 173 4 was be AUX cana-3719 173 5 performed perform VERB cana-3719 173 6 by by ADP cana-3719 173 7 a a DET cana-3719 173 8 large large ADJ cana-3719 173 9 receptive receptive ADJ cana-3719 173 10 field field NOUN cana-3719 173 11 - - PUNCT cana-3719 173 12 based base VERB cana-3719 173 13 attention attention NOUN cana-3719 173 14 process process NOUN cana-3719 173 15 . . PUNCT cana-3719 174 1 liu liu PROPN cana-3719 174 2 et et PROPN cana-3719 174 3 al al PROPN cana-3719 174 4 . . PUNCT cana-3719 175 1 [ [ X cana-3719 175 2 56 56 NUM cana-3719 175 3 ] ] PUNCT cana-3719 175 4 combined combine VERB cana-3719 175 5 the the DET cana-3719 175 6 pros pro NOUN cana-3719 175 7 of of ADP cana-3719 175 8 the the DET cana-3719 175 9 transformer transformer NOUN cana-3719 175 10 model model NOUN cana-3719 175 11 , , PUNCT cana-3719 175 12 and and CCONJ cana-3719 175 13 cnn cnn PROPN cana-3719 175 14 model model NOUN cana-3719 175 15 to to PART cana-3719 175 16 derive derive VERB cana-3719 175 17 a a DET cana-3719 175 18 bidirectional bidirectional ADJ cana-3719 175 19 stream stream NOUN cana-3719 175 20 network network NOUN cana-3719 175 21 . . PUNCT cana-3719 176 1 the the DET cana-3719 176 2 cnn cnn PROPN cana-3719 176 3 network network NOUN cana-3719 176 4 is be AUX cana-3719 176 5 utilized utilize VERB cana-3719 176 6 to to PART cana-3719 176 7 collect collect VERB cana-3719 176 8 more more ADJ cana-3719 176 9 local local ADJ cana-3719 176 10 information information NOUN cana-3719 176 11 , , PUNCT cana-3719 176 12 while while SCONJ cana-3719 176 13 the the DET cana-3719 176 14 transformer transformer NOUN cana-3719 176 15 is be AUX cana-3719 176 16 utilized utilize VERB cana-3719 176 17 to to PART cana-3719 176 18 collect collect VERB cana-3719 176 19 relevant relevant ADJ cana-3719 176 20 features feature NOUN cana-3719 176 21 throughout throughout ADP cana-3719 176 22 the the DET cana-3719 176 23 image image NOUN cana-3719 176 24 . . PUNCT cana-3719 177 1 the the DET cana-3719 177 2 cnn cnn PROPN cana-3719 177 3 , , PUNCT cana-3719 177 4 and and CCONJ cana-3719 177 5 transformer transformer NOUN cana-3719 177 6 model model NOUN cana-3719 177 7 are be AUX cana-3719 177 8 developed develop VERB cana-3719 177 9 on on ADP cana-3719 177 10 a a DET cana-3719 177 11 hierarchical hierarchical ADJ cana-3719 177 12 encoder encoder NOUN cana-3719 177 13 and and CCONJ cana-3719 177 14 decoder decoder NOUN cana-3719 177 15 architecture architecture NOUN cana-3719 177 16 that that PRON cana-3719 177 17 minimizes minimize VERB cana-3719 177 18 the the DET cana-3719 177 19 complexity complexity NOUN cana-3719 177 20 . . PUNCT cana-3719 178 1 the the DET cana-3719 178 2 degradation degradation NOUN cana-3719 178 3 of of ADP cana-3719 178 4 gan gan ADJ cana-3719 178 5 performance performance NOUN cana-3719 178 6 for for ADP cana-3719 178 7 large large ADJ cana-3719 178 8 missed miss VERB cana-3719 178 9 regions region NOUN cana-3719 178 10 is be AUX cana-3719 178 11 addressed address VERB cana-3719 178 12 by by ADP cana-3719 178 13 the the DET cana-3719 178 14 authors author NOUN cana-3719 178 15 chen chen PROPN cana-3719 178 16 et et PROPN cana-3719 178 17 al al PROPN cana-3719 178 18 . . PUNCT cana-3719 179 1 [ [ X cana-3719 179 2 57 57 NUM cana-3719 179 3 ] ] PUNCT cana-3719 179 4 by by ADP cana-3719 179 5 proposing propose VERB cana-3719 179 6 a a DET cana-3719 179 7 backbone backbone NOUN cana-3719 179 8 network network NOUN cana-3719 179 9 . . PUNCT cana-3719 180 1 the the DET cana-3719 180 2 backbone backbone NOUN cana-3719 180 3 network network NOUN cana-3719 180 4 uses use VERB cana-3719 180 5 a a DET cana-3719 180 6 nonpooling nonpoole VERB cana-3719 180 7 structure structure NOUN cana-3719 180 8 to to PART cana-3719 180 9 reconstruct reconstruct VERB cana-3719 180 10 the the DET cana-3719 180 11 texture texture NOUN cana-3719 180 12 descriptors descriptor NOUN cana-3719 180 13 . . PUNCT cana-3719 181 1 thus thus ADV cana-3719 181 2 the the DET cana-3719 181 3 network network NOUN cana-3719 181 4 has have VERB cana-3719 181 5 an an DET cana-3719 181 6 image image NOUN cana-3719 181 7 refinement refinement NOUN cana-3719 181 8 network network NOUN cana-3719 181 9 that that PRON cana-3719 181 10 follows follow VERB cana-3719 181 11 the the DET cana-3719 181 12 texture texture ADJ cana-3719 181 13 reconstruction reconstruction NOUN cana-3719 181 14 and and CCONJ cana-3719 181 15 structural structural ADJ cana-3719 181 16 reconstruction reconstruction NOUN cana-3719 181 17 network network NOUN cana-3719 181 18 as as SCONJ cana-3719 181 19 illustrated illustrate VERB cana-3719 181 20 in in ADP cana-3719 181 21 fig fig NOUN cana-3719 181 22 . . PUNCT cana-3719 182 1 7 7 X cana-3719 182 2 . . X cana-3719 182 3 fig fig NOUN cana-3719 182 4 . . PUNCT cana-3719 183 1 7 7 NUM cana-3719 183 2 : : PUNCT cana-3719 183 3 structure structure NOUN cana-3719 183 4 of of ADP cana-3719 183 5 the the DET cana-3719 183 6 texture texture ADJ cana-3719 183 7 and and CCONJ cana-3719 183 8 structural structural ADJ cana-3719 183 9 - - PUNCT cana-3719 183 10 based base VERB cana-3719 183 11 approach approach NOUN cana-3719 183 12 multi multi ADJ cana-3719 183 13 - - NOUN cana-3719 183 14 stage stage ADJ cana-3719 183 15 inpainting inpainte VERB cana-3719 183 16 schemes scheme NOUN cana-3719 183 17 are be AUX cana-3719 183 18 also also ADV cana-3719 183 19 commonly commonly ADV cana-3719 183 20 used use VERB cana-3719 183 21 in in ADP cana-3719 183 22 recent recent ADJ cana-3719 183 23 years year NOUN cana-3719 183 24 . . PUNCT cana-3719 184 1 usually usually ADV cana-3719 184 2 , , PUNCT cana-3719 184 3 the the DET cana-3719 184 4 number number NOUN cana-3719 184 5 of of ADP cana-3719 184 6 refined refined ADJ cana-3719 184 7 networks network NOUN cana-3719 184 8 is be AUX cana-3719 184 9 improved improve VERB cana-3719 184 10 to to PART cana-3719 184 11 enhance enhance VERB cana-3719 184 12 the the DET cana-3719 184 13 inpainting inpainting ADJ cana-3719 184 14 performance performance NOUN cana-3719 184 15 . . PUNCT cana-3719 185 1 but but CCONJ cana-3719 185 2 increasing increase VERB cana-3719 185 3 the the DET cana-3719 185 4 refined refined ADJ cana-3719 185 5 networks network NOUN cana-3719 185 6 also also ADV cana-3719 185 7 degrades degrade VERB cana-3719 185 8 the the DET cana-3719 185 9 image image NOUN cana-3719 185 10 due due ADP cana-3719 185 11 to to ADP cana-3719 185 12 the the DET cana-3719 185 13 presence presence NOUN cana-3719 185 14 of of ADP cana-3719 185 15 artifacts artifact NOUN cana-3719 185 16 , , PUNCT cana-3719 185 17 and and CCONJ cana-3719 185 18 blurring blurring NOUN cana-3719 185 19 effects effect NOUN cana-3719 185 20 . . PUNCT cana-3719 186 1 to to PART cana-3719 186 2 address address VERB cana-3719 186 3 this this DET cana-3719 186 4 issue issue NOUN cana-3719 186 5 a a DET cana-3719 186 6 perspective perspective ADJ cana-3719 186 7 field field NOUN cana-3719 186 8 structure structure NOUN cana-3719 186 9 [ [ X cana-3719 186 10 58 58 NUM cana-3719 186 11 ] ] PUNCT cana-3719 186 12 is be AUX cana-3719 186 13 introduced introduce VERB cana-3719 186 14 that that PRON cana-3719 186 15 uses use VERB cana-3719 186 16 a a DET cana-3719 186 17 u u NOUN cana-3719 186 18 - - NOUN cana-3719 186 19 network network NOUN cana-3719 186 20 with with ADP cana-3719 186 21 a a DET cana-3719 186 22 multi multi ADJ cana-3719 186 23 - - ADJ cana-3719 186 24 head head ADJ cana-3719 186 25 attention attention NOUN cana-3719 186 26 structure structure NOUN cana-3719 186 27 . . PUNCT cana-3719 187 1 five five NUM cana-3719 187 2 receptive receptive ADJ cana-3719 187 3 field field NOUN cana-3719 187 4 sub sub NOUN cana-3719 187 5 - - NOUN cana-3719 187 6 networks network NOUN cana-3719 187 7 are be AUX cana-3719 187 8 used use VERB cana-3719 187 9 in in ADP cana-3719 187 10 decreasing decrease VERB cana-3719 187 11 order order NOUN cana-3719 187 12 of of ADP cana-3719 187 13 receptive receptive ADJ cana-3719 187 14 field field NOUN cana-3719 187 15 perspective perspective NOUN cana-3719 187 16 . . PUNCT cana-3719 188 1 the the DET cana-3719 188 2 influence influence NOUN cana-3719 188 3 of of ADP cana-3719 188 4 local local ADJ cana-3719 188 5 pixels pixel NOUN cana-3719 188 6 in in ADP cana-3719 188 7 the the DET cana-3719 188 8 convolution convolution NOUN cana-3719 188 9 process process NOUN cana-3719 188 10 is be AUX cana-3719 188 11 minimized minimize VERB cana-3719 188 12 by by ADP cana-3719 188 13 the the DET cana-3719 188 14 usage usage NOUN cana-3719 188 15 of of ADP cana-3719 188 16 the the DET cana-3719 188 17 transconv transconv NOUN cana-3719 188 18 structure structure NOUN cana-3719 188 19 . . PUNCT cana-3719 189 1 an an DET cana-3719 189 2 adaptive adaptive ADJ cana-3719 189 3 feedback feedback NOUN cana-3719 189 4 process process NOUN cana-3719 189 5 is be AUX cana-3719 189 6 utilized utilize VERB cana-3719 189 7 to to PART cana-3719 189 8 derive derive VERB cana-3719 189 9 an an DET cana-3719 189 10 adaptive adaptive ADJ cana-3719 189 11 feedback feedback NOUN cana-3719 189 12 communications communication NOUN cana-3719 189 13 on on ADP cana-3719 189 14 applied apply VERB cana-3719 189 15 nonlinear nonlinear ADJ cana-3719 189 16 analysis analysis NOUN cana-3719 189 17 issn issn NOUN cana-3719 189 18 : : PUNCT cana-3719 189 19 1074 1074 NUM cana-3719 189 20 - - PUNCT cana-3719 189 21 133x 133x NUM cana-3719 189 22 vol vol NOUN cana-3719 189 23 32 32 NUM cana-3719 189 24 no no NOUN cana-3719 189 25 . . PUNCT cana-3719 190 1 8s 8s PROPN cana-3719 190 2 ( ( PUNCT cana-3719 190 3 2025 2025 NUM cana-3719 190 4 ) ) PUNCT cana-3719 190 5 584 584 NUM cana-3719 190 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 190 7 network network NOUN cana-3719 190 8 [ [ X cana-3719 190 9 59 59 NUM cana-3719 190 10 ] ] PUNCT cana-3719 190 11 . . PUNCT cana-3719 191 1 this this DET cana-3719 191 2 network network NOUN cana-3719 191 3 have have VERB cana-3719 191 4 the the DET cana-3719 191 5 ability ability NOUN cana-3719 191 6 to to PART cana-3719 191 7 derive derive VERB cana-3719 191 8 the the DET cana-3719 191 9 inpainting inpainting ADJ cana-3719 191 10 result result NOUN cana-3719 191 11 , , PUNCT cana-3719 191 12 and and CCONJ cana-3719 191 13 uncertainty uncertainty NOUN cana-3719 191 14 map map VERB cana-3719 191 15 simultaneously simultaneously ADV cana-3719 191 16 . . PUNCT cana-3719 192 1 as as SCONJ cana-3719 192 2 the the DET cana-3719 192 3 iteration iteration NOUN cana-3719 192 4 proceeds proceed VERB cana-3719 192 5 , , PUNCT cana-3719 192 6 the the DET cana-3719 192 7 uncertainty uncertainty NOUN cana-3719 192 8 is be AUX cana-3719 192 9 gradually gradually ADV cana-3719 192 10 reduced reduce VERB cana-3719 192 11 by by ADP cana-3719 192 12 the the DET cana-3719 192 13 feedback feedback NOUN cana-3719 192 14 network network NOUN cana-3719 192 15 , , PUNCT cana-3719 192 16 which which PRON cana-3719 192 17 also also ADV cana-3719 192 18 improves improve VERB cana-3719 192 19 the the DET cana-3719 192 20 inpainting inpainting ADJ cana-3719 192 21 result result NOUN cana-3719 192 22 . . PUNCT cana-3719 193 1 fig fig NOUN cana-3719 193 2 . . PUNCT cana-3719 194 1 8 8 NUM cana-3719 194 2 : : PUNCT cana-3719 194 3 feature feature NOUN cana-3719 194 4 fusion fusion NOUN cana-3719 194 5 approaches approach NOUN cana-3719 194 6 ( ( PUNCT cana-3719 194 7 a a DET cana-3719 194 8 ) ) PUNCT cana-3719 194 9 fusion fusion NOUN cana-3719 194 10 at at ADP cana-3719 194 11 transformer transformer NOUN cana-3719 194 12 stream stream NOUN cana-3719 194 13 ( ( PUNCT cana-3719 194 14 b b NOUN cana-3719 194 15 ) ) PUNCT cana-3719 194 16 fusion fusion NOUN cana-3719 194 17 at at ADP cana-3719 194 18 cnn cnn PROPN cana-3719 194 19 stream stream PROPN cana-3719 194 20 ( ( PUNCT cana-3719 194 21 c c NOUN cana-3719 194 22 ) ) PUNCT cana-3719 194 23 unified unified ADJ cana-3719 194 24 fusion fusion NOUN cana-3719 194 25 ( ( PUNCT cana-3719 194 26 d d NOUN cana-3719 194 27 ) ) PUNCT cana-3719 194 28 fusion fusion NOUN cana-3719 194 29 in in ADP cana-3719 194 30 both both DET cana-3719 194 31 transformer transformer NOUN cana-3719 194 32 and and CCONJ cana-3719 194 33 cnn cnn PROPN cana-3719 194 34 stream stream VERB cana-3719 194 35 the the DET cana-3719 194 36 features feature NOUN cana-3719 194 37 obtained obtain VERB cana-3719 194 38 by by ADP cana-3719 194 39 individual individual ADJ cana-3719 194 40 networks network NOUN cana-3719 194 41 are be AUX cana-3719 194 42 combined combine VERB cana-3719 194 43 by by ADP cana-3719 194 44 the the DET cana-3719 194 45 fusion fusion NOUN cana-3719 194 46 process process NOUN cana-3719 194 47 . . PUNCT cana-3719 195 1 i.e. i.e. X cana-3719 195 2 the the DET cana-3719 195 3 texture texture NOUN cana-3719 195 4 features feature NOUN cana-3719 195 5 obtained obtain VERB cana-3719 195 6 by by ADP cana-3719 195 7 one one NUM cana-3719 195 8 network network NOUN cana-3719 195 9 and and CCONJ cana-3719 195 10 the the DET cana-3719 195 11 structural structural ADJ cana-3719 195 12 feature feature NOUN cana-3719 195 13 obtained obtain VERB cana-3719 195 14 by by ADP cana-3719 195 15 another another DET cana-3719 195 16 network network NOUN cana-3719 195 17 are be AUX cana-3719 195 18 combined combine VERB cana-3719 195 19 by by ADP cana-3719 195 20 one one NUM cana-3719 195 21 of of ADP cana-3719 195 22 the the DET cana-3719 195 23 fusion fusion NOUN cana-3719 195 24 processes process NOUN cana-3719 195 25 illustrated illustrate VERB cana-3719 195 26 in in ADP cana-3719 195 27 fig fig NOUN cana-3719 195 28 . . PUNCT cana-3719 196 1 8 8 X cana-3719 196 2 . . X cana-3719 197 1 here here ADV cana-3719 197 2 four four NUM cana-3719 197 3 feature feature NOUN cana-3719 197 4 fusion fusion NOUN cana-3719 197 5 approaches approach NOUN cana-3719 197 6 are be AUX cana-3719 197 7 utilized utilize VERB cana-3719 197 8 to to PART cana-3719 197 9 merge merge VERB cana-3719 197 10 the the DET cana-3719 197 11 transformer transformer NOUN cana-3719 197 12 features feature NOUN cana-3719 197 13 ( ( PUNCT cana-3719 197 14 𝐺𝑑 𝐺𝑑 NOUN cana-3719 197 15 ) ) PUNCT cana-3719 197 16 and and CCONJ cana-3719 197 17 the the DET cana-3719 197 18 cnn cnn PROPN cana-3719 197 19 feature feature NOUN cana-3719 197 20 ( ( PUNCT cana-3719 197 21 𝐺𝑐 𝐺𝑐 PROPN cana-3719 197 22 ) ) PUNCT cana-3719 197 23 . . PUNCT cana-3719 198 1 4 4 X cana-3719 198 2 . . X cana-3719 198 3 conclusion conclusion NOUN cana-3719 198 4 this this DET cana-3719 198 5 paper paper NOUN cana-3719 198 6 delivers deliver VERB cana-3719 198 7 a a DET cana-3719 198 8 systematic systematic ADJ cana-3719 198 9 review review NOUN cana-3719 198 10 of of ADP cana-3719 198 11 different different ADJ cana-3719 198 12 traditional traditional ADJ cana-3719 198 13 and and CCONJ cana-3719 198 14 deep deep ADJ cana-3719 198 15 learning learning NOUN cana-3719 198 16 schemes scheme NOUN cana-3719 198 17 used use VERB cana-3719 198 18 for for ADP cana-3719 198 19 picture picture NOUN cana-3719 198 20 inpainting inpainte VERB cana-3719 198 21 applications application NOUN cana-3719 198 22 . . PUNCT cana-3719 199 1 the the DET cana-3719 199 2 paper paper NOUN cana-3719 199 3 discusses discuss VERB cana-3719 199 4 the the DET cana-3719 199 5 uniqueness uniqueness NOUN cana-3719 199 6 of of ADP cana-3719 199 7 the the DET cana-3719 199 8 different different ADJ cana-3719 199 9 research research NOUN cana-3719 199 10 works work NOUN cana-3719 199 11 that that PRON cana-3719 199 12 are be AUX cana-3719 199 13 derived derive VERB cana-3719 199 14 from from ADP cana-3719 199 15 exemplar exemplar ADJ cana-3719 199 16 , , PUNCT cana-3719 199 17 diffusion diffusion NOUN cana-3719 199 18 , , PUNCT cana-3719 199 19 and and CCONJ cana-3719 199 20 sparsity sparsity NOUN cana-3719 199 21 - - PUNCT cana-3719 199 22 based base VERB cana-3719 199 23 schemes scheme NOUN cana-3719 199 24 . . PUNCT cana-3719 200 1 the the DET cana-3719 200 2 work work NOUN cana-3719 200 3 also also ADV cana-3719 200 4 reviews review VERB cana-3719 200 5 the the DET cana-3719 200 6 various various ADJ cana-3719 200 7 deep deep ADJ cana-3719 200 8 learning learning NOUN cana-3719 200 9 structures structure NOUN cana-3719 200 10 that that PRON cana-3719 200 11 are be AUX cana-3719 200 12 used use VERB cana-3719 200 13 to to PART cana-3719 200 14 perform perform VERB cana-3719 200 15 image image NOUN cana-3719 200 16 inpainting inpainte VERB cana-3719 200 17 that that PRON cana-3719 200 18 effectively effectively ADV cana-3719 200 19 reconstructs reconstruct VERB cana-3719 200 20 the the DET cana-3719 200 21 texture texture NOUN cana-3719 200 22 and and CCONJ cana-3719 200 23 structure structure NOUN cana-3719 200 24 of of ADP cana-3719 200 25 the the DET cana-3719 200 26 missed miss VERB cana-3719 200 27 area area NOUN cana-3719 200 28 . . PUNCT cana-3719 201 1 more more ADV cana-3719 201 2 specifically specifically ADV cana-3719 201 3 the the DET cana-3719 201 4 paper paper NOUN cana-3719 201 5 also also ADV cana-3719 201 6 focuses focus VERB cana-3719 201 7 on on ADP cana-3719 201 8 the the DET cana-3719 201 9 hybrid hybrid ADJ cana-3719 201 10 architectures architecture NOUN cana-3719 201 11 that that PRON cana-3719 201 12 combine combine VERB cana-3719 201 13 the the DET cana-3719 201 14 advantage advantage NOUN cana-3719 201 15 of of ADP cana-3719 201 16 cnn cnn PROPN cana-3719 201 17 and and CCONJ cana-3719 201 18 gan gin VERB cana-3719 201 19 architecture architecture NOUN cana-3719 201 20 in in ADP cana-3719 201 21 the the DET cana-3719 201 22 reconstruction reconstruction NOUN cana-3719 201 23 of of ADP cana-3719 201 24 local local ADJ cana-3719 201 25 texture texture NOUN cana-3719 201 26 and and CCONJ cana-3719 201 27 global global ADJ cana-3719 201 28 structure structure NOUN cana-3719 201 29 . . PUNCT cana-3719 202 1 the the DET cana-3719 202 2 paper paper NOUN cana-3719 202 3 also also ADV cana-3719 202 4 reviews review VERB cana-3719 202 5 the the DET cana-3719 202 6 architectures architecture NOUN cana-3719 202 7 that that PRON cana-3719 202 8 are be AUX cana-3719 202 9 derived derive VERB cana-3719 202 10 from from ADP cana-3719 202 11 the the DET cana-3719 202 12 attention attention NOUN cana-3719 202 13 approaches approach NOUN cana-3719 202 14 , , PUNCT cana-3719 202 15 u u NOUN cana-3719 202 16 - - ADJ cana-3719 202 17 net net ADJ cana-3719 202 18 structures structure NOUN cana-3719 202 19 , , PUNCT cana-3719 202 20 and and CCONJ cana-3719 202 21 transformer transformer NOUN cana-3719 202 22 - - PUNCT cana-3719 202 23 based base VERB cana-3719 202 24 approaches approach NOUN cana-3719 202 25 that that PRON cana-3719 202 26 show show VERB cana-3719 202 27 better well ADJ cana-3719 202 28 performance performance NOUN cana-3719 202 29 for for ADP cana-3719 202 30 higher high ADJ cana-3719 202 31 mask mask NOUN cana-3719 202 32 ratios ratio NOUN cana-3719 202 33 . . PUNCT cana-3719 203 1 references reference NOUN cana-3719 203 2 [ [ X cana-3719 203 3 1 1 NUM cana-3719 203 4 ] ] PUNCT cana-3719 203 5 . . PUNCT cana-3719 204 1 bugeau bugeau PROPN cana-3719 204 2 , , PUNCT cana-3719 204 3 a. a. PROPN cana-3719 204 4 , , PUNCT cana-3719 204 5 bertalmΓ­o bertalmΓ­o PROPN cana-3719 204 6 , , PUNCT cana-3719 204 7 m. m. NOUN cana-3719 204 8 , , PUNCT cana-3719 204 9 caselles caselle NOUN cana-3719 204 10 , , PUNCT cana-3719 204 11 v. v. ADP cana-3719 204 12 , , PUNCT cana-3719 204 13 sapiro sapiro NOUN cana-3719 204 14 , , PUNCT cana-3719 204 15 g. g. PROPN cana-3719 204 16 ( ( PUNCT cana-3719 204 17 2010 2010 NUM cana-3719 204 18 ) ) PUNCT cana-3719 204 19 . . PUNCT cana-3719 205 1 a a DET cana-3719 205 2 comprehensive comprehensive ADJ cana-3719 205 3 framework framework NOUN cana-3719 205 4 for for ADP cana-3719 205 5 image image NOUN cana-3719 205 6 inpainting inpainte VERB cana-3719 205 7 . . PUNCT cana-3719 206 1 ieee ieee NOUN cana-3719 206 2 transactions transaction NOUN cana-3719 206 3 on on ADP cana-3719 206 4 image image NOUN cana-3719 206 5 processing processing NOUN cana-3719 206 6 , , PUNCT cana-3719 206 7 19(10 19(10 NUM cana-3719 206 8 ) ) PUNCT cana-3719 206 9 , , PUNCT cana-3719 206 10 2634 2634 NUM cana-3719 206 11 - - SYM cana-3719 206 12 2645 2645 NUM cana-3719 206 13 . . PUNCT cana-3719 207 1 [ [ X cana-3719 207 2 2 2 NUM cana-3719 207 3 ] ] X cana-3719 207 4 guillemot guillemot NOUN cana-3719 207 5 , , PUNCT cana-3719 207 6 c. c. PROPN cana-3719 207 7 , , PUNCT cana-3719 207 8 le le X cana-3719 207 9 meur meur NOUN cana-3719 207 10 , , PUNCT cana-3719 207 11 o. o. INTJ cana-3719 207 12 ( ( PUNCT cana-3719 207 13 2013 2013 NUM cana-3719 207 14 ) ) PUNCT cana-3719 207 15 . . PUNCT cana-3719 208 1 image image NOUN cana-3719 208 2 inpainting inpainte VERB cana-3719 208 3 : : PUNCT cana-3719 208 4 overview overview NOUN cana-3719 208 5 and and CCONJ cana-3719 208 6 recent recent ADJ cana-3719 208 7 advances advance NOUN cana-3719 208 8 . . PUNCT cana-3719 209 1 ieee ieee NOUN cana-3719 209 2 signal signal PROPN cana-3719 209 3 processing processing NOUN cana-3719 209 4 magazine magazine NOUN cana-3719 209 5 , , PUNCT cana-3719 209 6 31(1 31(1 NUM cana-3719 209 7 ) ) PUNCT cana-3719 209 8 , , PUNCT cana-3719 209 9 127 127 NUM cana-3719 209 10 - - SYM cana-3719 209 11 144 144 NUM cana-3719 209 12 . . PUNCT cana-3719 210 1 [ [ X cana-3719 210 2 3 3 NUM cana-3719 210 3 ] ] PUNCT cana-3719 210 4 . . PUNCT cana-3719 211 1 xu xu PROPN cana-3719 211 2 , , PUNCT cana-3719 211 3 z. z. PROPN cana-3719 211 4 , , PUNCT cana-3719 211 5 zhang zhang PROPN cana-3719 211 6 , , PUNCT cana-3719 211 7 x. x. PROPN cana-3719 211 8 , , PUNCT cana-3719 211 9 chen chen PROPN cana-3719 211 10 , , PUNCT cana-3719 211 11 w. w. PROPN cana-3719 211 12 , , PUNCT cana-3719 211 13 yao yao PROPN cana-3719 211 14 , , PUNCT cana-3719 211 15 m. m. NOUN cana-3719 211 16 , , PUNCT cana-3719 211 17 liu liu PROPN cana-3719 211 18 , , PUNCT cana-3719 211 19 j. j. PROPN cana-3719 211 20 , , PUNCT cana-3719 211 21 xu xu PROPN cana-3719 211 22 , , PUNCT cana-3719 211 23 t. t. PROPN cana-3719 211 24 , , PUNCT cana-3719 211 25 wang wang PROPN cana-3719 211 26 , , PUNCT cana-3719 211 27 z. z. PROPN cana-3719 211 28 ( ( PUNCT cana-3719 211 29 2023 2023 NUM cana-3719 211 30 ) ) PUNCT cana-3719 211 31 . . PUNCT cana-3719 212 1 a a DET cana-3719 212 2 review review NOUN cana-3719 212 3 of of ADP cana-3719 212 4 image image NOUN cana-3719 212 5 inpainting inpainte VERB cana-3719 212 6 methods method NOUN cana-3719 212 7 based base VERB cana-3719 212 8 on on ADP cana-3719 212 9 deep deep ADJ cana-3719 212 10 learning learning NOUN cana-3719 212 11 . . PUNCT cana-3719 213 1 applied apply VERB cana-3719 213 2 sciences science NOUN cana-3719 213 3 , , PUNCT cana-3719 213 4 13(20 13(20 NUM cana-3719 213 5 ) ) PUNCT cana-3719 213 6 , , PUNCT cana-3719 213 7 11189 11189 NUM cana-3719 213 8 . . PUNCT cana-3719 214 1 [ [ X cana-3719 214 2 4 4 NUM cana-3719 214 3 ] ] PUNCT cana-3719 214 4 . . PUNCT cana-3719 215 1 corneanu corneanu NOUN cana-3719 215 2 , , PUNCT cana-3719 215 3 ciprian ciprian PROPN cana-3719 215 4 , , PUNCT cana-3719 215 5 raghudeep raghudeep NOUN cana-3719 215 6 gadde gadde NOUN cana-3719 215 7 , , PUNCT cana-3719 215 8 and and CCONJ cana-3719 215 9 aleix aleix VERB cana-3719 215 10 m. m. PROPN cana-3719 215 11 martinez martinez PROPN cana-3719 215 12 . . PUNCT cana-3719 216 1 " " PUNCT cana-3719 216 2 latentpaint latentpaint NOUN cana-3719 216 3 : : PUNCT cana-3719 216 4 image image NOUN cana-3719 216 5 inpainting inpainte VERB cana-3719 216 6 in in ADP cana-3719 216 7 latent latent ADJ cana-3719 216 8 space space NOUN cana-3719 216 9 with with ADP cana-3719 216 10 diffusion diffusion NOUN cana-3719 216 11 models model NOUN cana-3719 216 12 . . PUNCT cana-3719 216 13 " " PUNCT cana-3719 217 1 proceedings proceeding NOUN cana-3719 217 2 of of ADP cana-3719 217 3 the the DET cana-3719 217 4 ieee ieee NOUN cana-3719 217 5 / / SYM cana-3719 217 6 cvf cvf NOUN cana-3719 217 7 winter winter NOUN cana-3719 217 8 conference conference NOUN cana-3719 217 9 on on ADP cana-3719 217 10 applications application NOUN cana-3719 217 11 of of ADP cana-3719 217 12 computer computer NOUN cana-3719 217 13 vision vision NOUN cana-3719 217 14 . . PUNCT cana-3719 218 1 2024 2024 NUM cana-3719 218 2 . . PUNCT cana-3719 219 1 [ [ X cana-3719 219 2 5 5 NUM cana-3719 219 3 ] ] PUNCT cana-3719 219 4 . . PUNCT cana-3719 219 5 arias arias PROPN cana-3719 219 6 , , PUNCT cana-3719 219 7 p. p. NOUN cana-3719 219 8 , , PUNCT cana-3719 219 9 facciolo facciolo PROPN cana-3719 219 10 , , PUNCT cana-3719 219 11 g. g. PROPN cana-3719 219 12 , , PUNCT cana-3719 219 13 caselles caselle NOUN cana-3719 219 14 , , PUNCT cana-3719 219 15 v. v. ADP cana-3719 219 16 , , PUNCT cana-3719 219 17 sapiro sapiro NOUN cana-3719 219 18 , , PUNCT cana-3719 219 19 g. g. PROPN cana-3719 219 20 ( ( PUNCT cana-3719 219 21 2011 2011 NUM cana-3719 219 22 ) ) PUNCT cana-3719 219 23 . . PUNCT cana-3719 220 1 a a DET cana-3719 220 2 variational variational ADJ cana-3719 220 3 framework framework NOUN cana-3719 220 4 for for ADP cana-3719 220 5 exemplar exemplar NOUN cana-3719 220 6 - - PUNCT cana-3719 220 7 based base VERB cana-3719 220 8 image image NOUN cana-3719 220 9 inpainting inpainte VERB cana-3719 220 10 . . PUNCT cana-3719 221 1 international international ADJ cana-3719 221 2 journal journal NOUN cana-3719 221 3 of of ADP cana-3719 221 4 computer computer NOUN cana-3719 221 5 vision vision NOUN cana-3719 221 6 , , PUNCT cana-3719 221 7 93 93 NUM cana-3719 221 8 , , PUNCT cana-3719 221 9 319 319 NUM cana-3719 221 10 - - SYM cana-3719 221 11 347 347 NUM cana-3719 221 12 . . PUNCT cana-3719 222 1 [ [ X cana-3719 222 2 6 6 NUM cana-3719 222 3 ] ] PUNCT cana-3719 222 4 . . PUNCT cana-3719 223 1 shen shen PROPN cana-3719 223 2 , , PUNCT cana-3719 223 3 b. b. PROPN cana-3719 223 4 , , PUNCT cana-3719 223 5 hu hu PROPN cana-3719 223 6 , , PUNCT cana-3719 223 7 w. w. PROPN cana-3719 223 8 , , PUNCT cana-3719 223 9 zhang zhang PROPN cana-3719 223 10 , , PUNCT cana-3719 223 11 y. y. PROPN cana-3719 223 12 , , PUNCT cana-3719 223 13 zhang zhang PROPN cana-3719 223 14 , , PUNCT cana-3719 223 15 y. y. PROPN cana-3719 223 16 j. j. PROPN cana-3719 223 17 ( ( PUNCT cana-3719 223 18 2009 2009 NUM cana-3719 223 19 , , PUNCT cana-3719 223 20 april april PROPN cana-3719 223 21 ) ) PUNCT cana-3719 223 22 . . PUNCT cana-3719 224 1 image image NOUN cana-3719 224 2 inpainting inpainte VERB cana-3719 224 3 via via ADP cana-3719 224 4 sparse sparse ADJ cana-3719 224 5 representation representation NOUN cana-3719 224 6 . . PUNCT cana-3719 225 1 in in ADP cana-3719 225 2 2009 2009 NUM cana-3719 225 3 ieee ieee NOUN cana-3719 225 4 international international ADJ cana-3719 225 5 conference conference NOUN cana-3719 225 6 on on ADP cana-3719 225 7 acoustics acoustic NOUN cana-3719 225 8 , , PUNCT cana-3719 225 9 speech speech NOUN cana-3719 225 10 and and CCONJ cana-3719 225 11 signal signal NOUN cana-3719 225 12 processing processing NOUN cana-3719 225 13 ( ( PUNCT cana-3719 225 14 pp pp ADJ cana-3719 225 15 . . PUNCT cana-3719 226 1 697 697 NUM cana-3719 226 2 - - NUM cana-3719 226 3 700 700 NUM cana-3719 226 4 ) ) PUNCT cana-3719 226 5 . . PUNCT cana-3719 227 1 ieee ieee NOUN cana-3719 227 2 . . PUNCT cana-3719 228 1 [ [ X cana-3719 228 2 7 7 NUM cana-3719 228 3 ] ] PUNCT cana-3719 228 4 . . PUNCT cana-3719 229 1 li li PROPN cana-3719 229 2 , , PUNCT cana-3719 229 3 z. z. PROPN cana-3719 229 4 , , PUNCT cana-3719 229 5 wu wu PROPN cana-3719 229 6 , , PUNCT cana-3719 229 7 j. j. PROPN cana-3719 229 8 ( ( PUNCT cana-3719 229 9 2019 2019 NUM cana-3719 229 10 ) ) PUNCT cana-3719 229 11 . . PUNCT cana-3719 230 1 learning learn VERB cana-3719 230 2 deep deep ADJ cana-3719 230 3 cnn cnn PROPN cana-3719 230 4 denoiser denoiser PROPN cana-3719 230 5 priors prior NOUN cana-3719 230 6 for for ADP cana-3719 230 7 depth depth NOUN cana-3719 230 8 image image NOUN cana-3719 230 9 inpainting inpainte VERB cana-3719 230 10 . . PUNCT cana-3719 231 1 applied apply VERB cana-3719 231 2 sciences science NOUN cana-3719 231 3 , , PUNCT cana-3719 231 4 9(6 9(6 NUM cana-3719 231 5 ) ) PUNCT cana-3719 231 6 , , PUNCT cana-3719 231 7 1103 1103 NUM cana-3719 231 8 . . PUNCT cana-3719 232 1 [ [ X cana-3719 232 2 8 8 NUM cana-3719 232 3 ] ] PUNCT cana-3719 232 4 . . PUNCT cana-3719 233 1 zhang zhang PROPN cana-3719 233 2 , , PUNCT cana-3719 233 3 x. x. PROPN cana-3719 233 4 , , PUNCT cana-3719 233 5 wang wang PROPN cana-3719 233 6 , , PUNCT cana-3719 233 7 x. x. PROPN cana-3719 233 8 , , PUNCT cana-3719 233 9 shi shi PROPN cana-3719 233 10 , , PUNCT cana-3719 233 11 c. c. PROPN cana-3719 233 12 , , PUNCT cana-3719 233 13 yan yan PROPN cana-3719 233 14 , , PUNCT cana-3719 233 15 z. z. PROPN cana-3719 233 16 , , PUNCT cana-3719 233 17 li li PROPN cana-3719 233 18 , , PUNCT cana-3719 233 19 x. x. PROPN cana-3719 233 20 , , PUNCT cana-3719 233 21 kong kong PROPN cana-3719 233 22 , , PUNCT cana-3719 233 23 b. b. PROPN cana-3719 233 24 , , PUNCT cana-3719 233 25 ... ... PUNCT cana-3719 233 26 mumtaz mumtaz PROPN cana-3719 233 27 , , PUNCT cana-3719 233 28 i. i. PROPN cana-3719 233 29 ( ( PUNCT cana-3719 233 30 2022 2022 NUM cana-3719 233 31 ) ) PUNCT cana-3719 233 32 . . PUNCT cana-3719 234 1 de de PROPN cana-3719 234 2 - - ADJ cana-3719 234 3 gan gan ADJ cana-3719 234 4 : : PUNCT cana-3719 234 5 domain domain NOUN cana-3719 234 6 embedded embed VERB cana-3719 234 7 gan gan NOUN cana-3719 234 8 for for ADP cana-3719 234 9 high high ADJ cana-3719 234 10 quality quality NOUN cana-3719 234 11 face face NOUN cana-3719 234 12 image image NOUN cana-3719 234 13 inpainting inpainte VERB cana-3719 234 14 . . PUNCT cana-3719 235 1 pattern pattern NOUN cana-3719 235 2 recognition recognition NOUN cana-3719 235 3 , , PUNCT cana-3719 235 4 124 124 NUM cana-3719 235 5 , , PUNCT cana-3719 235 6 108415 108415 NUM cana-3719 235 7 . . PUNCT cana-3719 236 1 [ [ X cana-3719 236 2 9 9 NUM cana-3719 236 3 ] ] PUNCT cana-3719 236 4 . . PUNCT cana-3719 237 1 ko ko PROPN cana-3719 237 2 , , PUNCT cana-3719 237 3 k. k. PROPN cana-3719 237 4 , , PUNCT cana-3719 237 5 kim kim PROPN cana-3719 237 6 , , PUNCT cana-3719 237 7 c. c. PROPN cana-3719 237 8 s. s. PROPN cana-3719 237 9 ( ( PUNCT cana-3719 237 10 2023 2023 NUM cana-3719 237 11 ) ) PUNCT cana-3719 237 12 . . PUNCT cana-3719 238 1 continuously continuously ADV cana-3719 238 2 masked mask VERB cana-3719 238 3 transformer transformer NOUN cana-3719 238 4 for for ADP cana-3719 238 5 image image NOUN cana-3719 238 6 inpainting inpainte VERB cana-3719 238 7 . . PUNCT cana-3719 239 1 in in ADP cana-3719 239 2 proceedings proceeding NOUN cana-3719 239 3 of of ADP cana-3719 239 4 the the DET cana-3719 239 5 ieee ieee NOUN cana-3719 239 6 / / SYM cana-3719 239 7 cvf cvf NOUN cana-3719 239 8 international international ADJ cana-3719 239 9 conference conference NOUN cana-3719 239 10 on on ADP cana-3719 239 11 computer computer NOUN cana-3719 239 12 vision vision NOUN cana-3719 239 13 ( ( PUNCT cana-3719 239 14 pp pp ADJ cana-3719 239 15 . . PUNCT cana-3719 239 16 13169 13169 NUM cana-3719 239 17 - - SYM cana-3719 239 18 13178 13178 NUM cana-3719 239 19 ) ) PUNCT cana-3719 239 20 . . PUNCT cana-3719 240 1 [ [ X cana-3719 240 2 10 10 NUM cana-3719 240 3 ] ] PUNCT cana-3719 240 4 . . PUNCT cana-3719 241 1 wei wei PROPN cana-3719 241 2 , , PUNCT cana-3719 241 3 r. r. PROPN cana-3719 241 4 , , PUNCT cana-3719 241 5 wu wu PROPN cana-3719 241 6 , , PUNCT cana-3719 241 7 y. y. PROPN cana-3719 241 8 ( ( PUNCT cana-3719 241 9 2022 2022 NUM cana-3719 241 10 ) ) PUNCT cana-3719 241 11 . . PUNCT cana-3719 242 1 image image NOUN cana-3719 242 2 inpainting inpainte VERB cana-3719 242 3 via via ADP cana-3719 242 4 context context PROPN cana-3719 242 5 discriminator discriminator NOUN cana-3719 242 6 and and CCONJ cana-3719 242 7 u‐net u‐net PROPN cana-3719 242 8 . . PUNCT cana-3719 243 1 mathematical mathematical ADJ cana-3719 243 2 problems problem NOUN cana-3719 243 3 in in ADP cana-3719 243 4 engineering engineering NOUN cana-3719 243 5 , , PUNCT cana-3719 243 6 2022(1 2022(1 NUM cana-3719 243 7 ) ) PUNCT cana-3719 243 8 , , PUNCT cana-3719 243 9 7328045 7328045 NUM cana-3719 243 10 . . PUNCT cana-3719 244 1 [ [ X cana-3719 244 2 11 11 NUM cana-3719 244 3 ] ] PUNCT cana-3719 244 4 . . PUNCT cana-3719 245 1 sridevi sridevi PROPN cana-3719 245 2 , , PUNCT cana-3719 245 3 g. g. PROPN cana-3719 245 4 , , PUNCT cana-3719 245 5 srinivas srinivas PROPN cana-3719 245 6 kumar kumar PROPN cana-3719 245 7 , , PUNCT cana-3719 245 8 s. s. PROPN cana-3719 245 9 ( ( PUNCT cana-3719 245 10 2019 2019 NUM cana-3719 245 11 ) ) PUNCT cana-3719 245 12 . . PUNCT cana-3719 246 1 image image NOUN cana-3719 246 2 inpainting inpainte VERB cana-3719 246 3 based base VERB cana-3719 246 4 on on ADP cana-3719 246 5 fractional fractional ADJ cana-3719 246 6 -order -order PROPN cana-3719 246 7 nonlinear nonlinear ADJ cana-3719 246 8 diffusion diffusion NOUN cana-3719 246 9 for for ADP cana-3719 246 10 image image NOUN cana-3719 246 11 reconstruction reconstruction NOUN cana-3719 246 12 . . PUNCT cana-3719 247 1 circuits circuit NOUN cana-3719 247 2 , , PUNCT cana-3719 247 3 systems system NOUN cana-3719 247 4 , , PUNCT cana-3719 247 5 and and CCONJ cana-3719 247 6 signal signal NOUN cana-3719 247 7 processing processing NOUN cana-3719 247 8 , , PUNCT cana-3719 247 9 38 38 NUM cana-3719 247 10 , , PUNCT cana-3719 247 11 3802 3802 NUM cana-3719 247 12 - - SYM cana-3719 247 13 3817 3817 NUM cana-3719 247 14 . . PUNCT cana-3719 248 1 communications communication NOUN cana-3719 248 2 on on ADP cana-3719 248 3 applied apply VERB cana-3719 248 4 nonlinear nonlinear ADJ cana-3719 248 5 analysis analysis NOUN cana-3719 248 6 issn issn NOUN cana-3719 248 7 : : PUNCT cana-3719 248 8 1074 1074 NUM cana-3719 248 9 - - PUNCT cana-3719 248 10 133x 133x NUM cana-3719 248 11 vol vol NOUN cana-3719 248 12 32 32 NUM cana-3719 248 13 no no NOUN cana-3719 248 14 . . PUNCT cana-3719 249 1 8s 8s PROPN cana-3719 249 2 ( ( PUNCT cana-3719 249 3 2025 2025 NUM cana-3719 249 4 ) ) PUNCT cana-3719 249 5 585 585 NUM cana-3719 249 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 250 1 [ [ X cana-3719 250 2 12 12 NUM cana-3719 250 3 ] ] PUNCT cana-3719 250 4 . . PUNCT cana-3719 251 1 j. j. PROPN cana-3719 251 2 song song PROPN cana-3719 251 3 , , PUNCT cana-3719 251 4 c. c. PROPN cana-3719 251 5 meng meng PROPN cana-3719 251 6 , , PUNCT cana-3719 251 7 and and CCONJ cana-3719 251 8 s. s. PROPN cana-3719 251 9 ermon ermon PROPN cana-3719 251 10 , , PUNCT cana-3719 251 11 β€˜ ' PUNCT cana-3719 251 12 β€˜ ' PUNCT cana-3719 251 13 denoising denoise VERB cana-3719 251 14 diffusion diffusion NOUN cana-3719 251 15 implicit implicit ADJ cana-3719 251 16 models model NOUN cana-3719 251 17 , , PUNCT cana-3719 251 18 ’’ ’' PUNCT cana-3719 251 19 2020 2020 NUM cana-3719 251 20 , , PUNCT cana-3719 251 21 arxiv:2010.02502 arxiv:2010.02502 PROPN cana-3719 251 22 . . PUNCT cana-3719 252 1 [ [ X cana-3719 252 2 13 13 NUM cana-3719 252 3 ] ] PUNCT cana-3719 252 4 . . PUNCT cana-3719 253 1 j. j. PROPN cana-3719 253 2 ho ho PROPN cana-3719 253 3 , , PUNCT cana-3719 253 4 a. a. NOUN cana-3719 253 5 jain jain PROPN cana-3719 253 6 , , PUNCT cana-3719 253 7 and and CCONJ cana-3719 253 8 p. p. PROPN cana-3719 253 9 abbeel abbeel NOUN cana-3719 253 10 , , PUNCT cana-3719 253 11 β€˜ ' PUNCT cana-3719 253 12 β€˜ ' PUNCT cana-3719 253 13 denoising denoise VERB cana-3719 253 14 diffusion diffusion NOUN cana-3719 253 15 probabilistic probabilistic ADJ cana-3719 253 16 models model NOUN cana-3719 253 17 , , PUNCT cana-3719 253 18 ’’ ’' PUNCT cana-3719 253 19 in in ADP cana-3719 253 20 proc proc NOUN cana-3719 253 21 . . PUNCT cana-3719 254 1 adv adv PROPN cana-3719 254 2 . . PUNCT cana-3719 254 3 neural neural PROPN cana-3719 254 4 inf inf PROPN cana-3719 254 5 . . PUNCT cana-3719 254 6 process process NOUN cana-3719 254 7 . . PUNCT cana-3719 255 1 syst syst PROPN cana-3719 255 2 . . PUNCT cana-3719 255 3 , , PUNCT cana-3719 255 4 vol vol NOUN cana-3719 255 5 . . PROPN cana-3719 255 6 33 33 NUM cana-3719 255 7 , , PUNCT cana-3719 255 8 2020 2020 NUM cana-3719 255 9 , , PUNCT cana-3719 255 10 pp pp ADJ cana-3719 255 11 . . PUNCT cana-3719 256 1 6840–6851 6840–6851 NUM cana-3719 256 2 . . PUNCT cana-3719 257 1 [ [ X cana-3719 257 2 14 14 NUM cana-3719 257 3 ] ] SYM cana-3719 257 4 . . PUNCT cana-3719 258 1 zhang zhang PROPN cana-3719 258 2 , , PUNCT cana-3719 258 3 n. n. PROPN cana-3719 258 4 , , PUNCT cana-3719 258 5 ji ji PROPN cana-3719 258 6 , , PUNCT cana-3719 258 7 h. h. PROPN cana-3719 258 8 , , PUNCT cana-3719 258 9 liu liu PROPN cana-3719 258 10 , , PUNCT cana-3719 258 11 l. l. PROPN cana-3719 258 12 , , PUNCT cana-3719 258 13 wang wang PROPN cana-3719 258 14 , , PUNCT cana-3719 258 15 g. g. PROPN cana-3719 258 16 ( ( PUNCT cana-3719 258 17 2019 2019 NUM cana-3719 258 18 ) ) PUNCT cana-3719 258 19 . . PUNCT cana-3719 259 1 exemplar exemplar NOUN cana-3719 259 2 - - PUNCT cana-3719 259 3 based base VERB cana-3719 259 4 image image NOUN cana-3719 259 5 inpainting inpainte VERB cana-3719 259 6 using use VERB cana-3719 259 7 angle angle NOUN cana-3719 259 8 - - PUNCT cana-3719 259 9 aware aware ADJ cana-3719 259 10 patch patch NOUN cana-3719 259 11 matching matching NOUN cana-3719 259 12 . . PUNCT cana-3719 260 1 eurasip eurasip PROPN cana-3719 260 2 journal journal PROPN cana-3719 260 3 on on ADP cana-3719 260 4 image image NOUN cana-3719 260 5 and and CCONJ cana-3719 260 6 video video NOUN cana-3719 260 7 processing processing NOUN cana-3719 260 8 , , PUNCT cana-3719 260 9 2019 2019 NUM cana-3719 260 10 , , PUNCT cana-3719 260 11 1 1 NUM cana-3719 260 12 - - SYM cana-3719 260 13 13 13 NUM cana-3719 260 14 . . PUNCT cana-3719 261 1 [ [ X cana-3719 261 2 15 15 NUM cana-3719 261 3 ] ] PUNCT cana-3719 261 4 . . PUNCT cana-3719 262 1 xu xu PROPN cana-3719 262 2 , , PUNCT cana-3719 262 3 z. z. PROPN cana-3719 262 4 , , PUNCT cana-3719 262 5 sun sun PROPN cana-3719 262 6 , , PUNCT cana-3719 262 7 j. j. PROPN cana-3719 262 8 ( ( PUNCT cana-3719 262 9 2010 2010 NUM cana-3719 262 10 ) ) PUNCT cana-3719 262 11 . . PUNCT cana-3719 263 1 image image NOUN cana-3719 263 2 inpainting inpainte VERB cana-3719 263 3 by by ADP cana-3719 263 4 patch patch ADJ cana-3719 263 5 propagation propagation NOUN cana-3719 263 6 using use VERB cana-3719 263 7 patch patch NOUN cana-3719 263 8 sparsity sparsity NOUN cana-3719 263 9 . . PUNCT cana-3719 264 1 ieee ieee NOUN cana-3719 264 2 transactions transaction NOUN cana-3719 264 3 on on ADP cana-3719 264 4 image image NOUN cana-3719 264 5 processing processing NOUN cana-3719 264 6 , , PUNCT cana-3719 264 7 19(5 19(5 NUM cana-3719 264 8 ) ) PUNCT cana-3719 264 9 , , PUNCT cana-3719 264 10 1153 1153 NUM cana-3719 264 11 - - SYM cana-3719 264 12 1165 1165 NUM cana-3719 264 13 . . PUNCT cana-3719 265 1 [ [ X cana-3719 265 2 16 16 NUM cana-3719 265 3 ] ] PUNCT cana-3719 265 4 . . PUNCT cana-3719 266 1 yi yi PROPN cana-3719 266 2 , , PUNCT cana-3719 266 3 z. z. PROPN cana-3719 266 4 , , PUNCT cana-3719 266 5 tang tang PROPN cana-3719 266 6 , , PUNCT cana-3719 266 7 q. q. PROPN cana-3719 266 8 , , PUNCT cana-3719 266 9 azizi azizi PROPN cana-3719 266 10 , , PUNCT cana-3719 266 11 s. s. PROPN cana-3719 266 12 , , PUNCT cana-3719 266 13 jang jang PROPN cana-3719 266 14 , , PUNCT cana-3719 266 15 d. d. PROPN cana-3719 266 16 , , PUNCT cana-3719 266 17 xu xu PROPN cana-3719 266 18 , , PUNCT cana-3719 266 19 z. z. PROPN cana-3719 266 20 ( ( PUNCT cana-3719 266 21 2020 2020 NUM cana-3719 266 22 ) ) PUNCT cana-3719 266 23 . . PUNCT cana-3719 267 1 contextual contextual ADJ cana-3719 267 2 residual residual ADJ cana-3719 267 3 aggregation aggregation NOUN cana-3719 267 4 for for ADP cana-3719 267 5 ultra ultra ADJ cana-3719 267 6 high high ADJ cana-3719 267 7 -resolution -resolution NOUN cana-3719 267 8 image image NOUN cana-3719 267 9 inpainting inpainte VERB cana-3719 267 10 . . PUNCT cana-3719 268 1 in in ADP cana-3719 268 2 proceedings proceeding NOUN cana-3719 268 3 of of ADP cana-3719 268 4 the the DET cana-3719 268 5 ieee ieee NOUN cana-3719 268 6 / / SYM cana-3719 268 7 cvf cvf NOUN cana-3719 268 8 conference conference NOUN cana-3719 268 9 on on ADP cana-3719 268 10 computer computer NOUN cana-3719 268 11 vision vision NOUN cana-3719 268 12 and and CCONJ cana-3719 268 13 pattern pattern NOUN cana-3719 268 14 recognition recognition NOUN cana-3719 268 15 ( ( PUNCT cana-3719 268 16 pp pp ADJ cana-3719 268 17 . . PUNCT cana-3719 268 18 7508 7508 NUM cana-3719 268 19 - - SYM cana-3719 268 20 7517 7517 NUM cana-3719 268 21 ) ) PUNCT cana-3719 268 22 . . PUNCT cana-3719 269 1 [ [ X cana-3719 269 2 17 17 NUM cana-3719 269 3 ] ] PUNCT cana-3719 269 4 . . PUNCT cana-3719 270 1 prados prados PROPN cana-3719 270 2 , , PUNCT cana-3719 270 3 r. r. PROPN cana-3719 270 4 , , PUNCT cana-3719 270 5 garcia garcia PROPN cana-3719 270 6 , , PUNCT cana-3719 270 7 r. r. PROPN cana-3719 270 8 , , PUNCT cana-3719 270 9 neumann neumann PROPN cana-3719 270 10 , , PUNCT cana-3719 270 11 l. l. PROPN cana-3719 270 12 ( ( PUNCT cana-3719 270 13 2014 2014 NUM cana-3719 270 14 ) ) PUNCT cana-3719 270 15 . . PUNCT cana-3719 271 1 image image NOUN cana-3719 271 2 blending blend VERB cana-3719 271 3 techniques technique NOUN cana-3719 271 4 and and CCONJ cana-3719 271 5 their their PRON cana-3719 271 6 application application NOUN cana-3719 271 7 in in ADP cana-3719 271 8 underwater underwater ADJ cana-3719 271 9 mosaicing mosaicing NOUN cana-3719 271 10 ( ( PUNCT cana-3719 271 11 vol vol NOUN cana-3719 271 12 . . PROPN cana-3719 271 13 13 13 NUM cana-3719 271 14 ) ) PUNCT cana-3719 271 15 . . PUNCT cana-3719 272 1 cham cham PROPN cana-3719 272 2 , , PUNCT cana-3719 272 3 switzerland switzerland PROPN cana-3719 272 4 : : PUNCT cana-3719 272 5 springer springer NOUN cana-3719 272 6 . . PUNCT cana-3719 273 1 [ [ X cana-3719 273 2 18 18 NUM cana-3719 273 3 ] ] PUNCT cana-3719 273 4 . . PUNCT cana-3719 274 1 ghrabat ghrabat PROPN cana-3719 274 2 , , PUNCT cana-3719 274 3 m. m. NOUN cana-3719 274 4 j. j. PROPN cana-3719 274 5 j. j. PROPN cana-3719 274 6 , , PUNCT cana-3719 274 7 ma ma PROPN cana-3719 274 8 , , PUNCT cana-3719 274 9 g. g. PROPN cana-3719 274 10 , , PUNCT cana-3719 274 11 abduljabbar abduljabbar PROPN cana-3719 274 12 , , PUNCT cana-3719 274 13 z. z. PROPN cana-3719 274 14 a. a. PROPN cana-3719 274 15 , , PUNCT cana-3719 274 16 al al PROPN cana-3719 274 17 sibahee sibahee PROPN cana-3719 274 18 , , PUNCT cana-3719 274 19 m. m. NOUN cana-3719 274 20 a. a. PROPN cana-3719 274 21 , , PUNCT cana-3719 274 22 jassim jassim PROPN cana-3719 274 23 , , PUNCT cana-3719 274 24 s. s. PROPN cana-3719 274 25 j. j. PROPN cana-3719 274 26 ( ( PUNCT cana-3719 274 27 2019 2019 NUM cana-3719 274 28 ) ) PUNCT cana-3719 274 29 . . PUNCT cana-3719 275 1 greedy greedy ADJ cana-3719 275 2 learning learning NOUN cana-3719 275 3 of of ADP cana-3719 275 4 deep deep ADJ cana-3719 275 5 boltzmann boltzmann PROPN cana-3719 275 6 machine machine NOUN cana-3719 275 7 ( ( PUNCT cana-3719 275 8 gdbm gdbm PROPN cana-3719 275 9 ) ) PUNCT cana-3719 275 10 ’s ’s PART cana-3719 275 11 variance variance NOUN cana-3719 275 12 and and CCONJ cana-3719 275 13 search search NOUN cana-3719 275 14 algorithm algorithm NOUN cana-3719 275 15 for for ADP cana-3719 275 16 efficient efficient ADJ cana-3719 275 17 image image NOUN cana-3719 275 18 retrieval retrieval NOUN cana-3719 275 19 . . PUNCT cana-3719 276 1 ieee ieee NOUN cana-3719 276 2 access access NOUN cana-3719 276 3 , , PUNCT cana-3719 276 4 7 7 NUM cana-3719 276 5 , , PUNCT cana-3719 276 6 169142169159 169142169159 NUM cana-3719 276 7 . . PUNCT cana-3719 277 1 [ [ X cana-3719 277 2 19 19 NUM cana-3719 277 3 ] ] PUNCT cana-3719 277 4 . . PUNCT cana-3719 278 1 li li PROPN cana-3719 278 2 , , PUNCT cana-3719 278 3 fan fan PROPN cana-3719 278 4 , , PUNCT cana-3719 278 5 et et PROPN cana-3719 278 6 al al PROPN cana-3719 278 7 . . PUNCT cana-3719 279 1 " " PUNCT cana-3719 279 2 magiceraser magiceraser NOUN cana-3719 279 3 : : PUNCT cana-3719 279 4 erasing erase VERB cana-3719 279 5 any any DET cana-3719 279 6 objects object NOUN cana-3719 279 7 via via ADP cana-3719 279 8 semantics semantic NOUN cana-3719 279 9 - - PUNCT cana-3719 279 10 aware aware ADJ cana-3719 279 11 control control NOUN cana-3719 279 12 . . PUNCT cana-3719 279 13 " " PUNCT cana-3719 280 1 european european ADJ cana-3719 280 2 conference conference NOUN cana-3719 280 3 on on ADP cana-3719 280 4 computer computer NOUN cana-3719 280 5 vision vision NOUN cana-3719 280 6 . . PUNCT cana-3719 281 1 springer springer NOUN cana-3719 281 2 , , PUNCT cana-3719 281 3 cham cham PROPN cana-3719 281 4 , , PUNCT cana-3719 281 5 2025 2025 NUM cana-3719 281 6 . . PUNCT cana-3719 282 1 [ [ X cana-3719 282 2 20 20 NUM cana-3719 282 3 ] ] PUNCT cana-3719 282 4 . . PUNCT cana-3719 283 1 han han PROPN cana-3719 283 2 , , PUNCT cana-3719 283 3 ruyi ruyi ADJ cana-3719 283 4 , , PUNCT cana-3719 283 5 et et PROPN cana-3719 283 6 al al PROPN cana-3719 283 7 . . PUNCT cana-3719 284 1 " " PUNCT cana-3719 284 2 transformed transform VERB cana-3719 284 3 sparsity sparsity NOUN cana-3719 284 4 - - PUNCT cana-3719 284 5 boosted boost VERB cana-3719 284 6 low low ADJ cana-3719 284 7 - - PUNCT cana-3719 284 8 rank rank NOUN cana-3719 284 9 model model NOUN cana-3719 284 10 for for ADP cana-3719 284 11 image image NOUN cana-3719 284 12 inpainting inpainte VERB cana-3719 284 13 with with ADP cana-3719 284 14 non non ADJ cana-3719 284 15 - - ADJ cana-3719 284 16 convex convex ADJ cana-3719 284 17 Ξ³ Ξ³ ADJ cana-3719 284 18 - - ADJ cana-3719 284 19 norm norm NOUN cana-3719 284 20 regularization regularization NOUN cana-3719 284 21 and and CCONJ cana-3719 284 22 non non ADJ cana-3719 284 23 - - ADJ cana-3719 284 24 local local ADJ cana-3719 284 25 prior prior NOUN cana-3719 284 26 . . PUNCT cana-3719 284 27 " " PUNCT cana-3719 285 1 optics optics PROPN cana-3719 285 2 & & CCONJ cana-3719 285 3 laser laser PROPN cana-3719 285 4 technology technology NOUN cana-3719 285 5 181 181 NUM cana-3719 285 6 ( ( PUNCT cana-3719 285 7 2025 2025 NUM cana-3719 285 8 ): ): PUNCT cana-3719 285 9 111865 111865 NUM cana-3719 285 10 . . PUNCT cana-3719 286 1 [ [ X cana-3719 286 2 21 21 NUM cana-3719 286 3 ] ] PUNCT cana-3719 286 4 . . PUNCT cana-3719 287 1 li li PROPN cana-3719 287 2 , , PUNCT cana-3719 287 3 z. z. PROPN cana-3719 287 4 , , PUNCT cana-3719 287 5 wu wu PROPN cana-3719 287 6 , , PUNCT cana-3719 287 7 j. j. PROPN cana-3719 287 8 ( ( PUNCT cana-3719 287 9 2019 2019 NUM cana-3719 287 10 ) ) PUNCT cana-3719 287 11 . . PUNCT cana-3719 288 1 learning learn VERB cana-3719 288 2 deep deep ADJ cana-3719 288 3 cnn cnn PROPN cana-3719 288 4 denoiser denoiser PROPN cana-3719 288 5 priors prior NOUN cana-3719 288 6 for for ADP cana-3719 288 7 depth depth NOUN cana-3719 288 8 image image NOUN cana-3719 288 9 inpainting inpainte VERB cana-3719 288 10 . . PUNCT cana-3719 289 1 applied apply VERB cana-3719 289 2 sciences science NOUN cana-3719 289 3 , , PUNCT cana-3719 289 4 9(6 9(6 NUM cana-3719 289 5 ) ) PUNCT cana-3719 289 6 , , PUNCT cana-3719 289 7 1103 1103 NUM cana-3719 289 8 . . PUNCT cana-3719 290 1 [ [ X cana-3719 290 2 22 22 NUM cana-3719 290 3 ] ] PUNCT cana-3719 290 4 . . PUNCT cana-3719 291 1 miao miao PROPN cana-3719 291 2 , , PUNCT cana-3719 291 3 j. j. PROPN cana-3719 291 4 , , PUNCT cana-3719 291 5 kou kou PROPN cana-3719 291 6 , , PUNCT cana-3719 291 7 k. k. PROPN cana-3719 291 8 i. i. PROPN cana-3719 291 9 , , PUNCT cana-3719 291 10 yang yang PROPN cana-3719 291 11 , , PUNCT cana-3719 291 12 y. y. PROPN cana-3719 291 13 , , PUNCT cana-3719 291 14 yang yang PROPN cana-3719 291 15 , , PUNCT cana-3719 291 16 l. l. PROPN cana-3719 291 17 , , PUNCT cana-3719 291 18 han han PROPN cana-3719 291 19 , , PUNCT cana-3719 291 20 j. j. PROPN cana-3719 291 21 ( ( PUNCT cana-3719 291 22 2024 2024 NUM cana-3719 291 23 ) ) PUNCT cana-3719 291 24 . . PUNCT cana-3719 292 1 quaternion quaternion NOUN cana-3719 292 2 matrix matrix NOUN cana-3719 292 3 completion completion NOUN cana-3719 292 4 using use VERB cana-3719 292 5 untrained untrained ADJ cana-3719 292 6 quaternion quaternion ADJ cana-3719 292 7 convolutional convolutional ADJ cana-3719 292 8 neural neural ADJ cana-3719 292 9 network network NOUN cana-3719 292 10 for for ADP cana-3719 292 11 color color NOUN cana-3719 292 12 image image NOUN cana-3719 292 13 inpainting inpainte VERB cana-3719 292 14 . . PUNCT cana-3719 292 15 signal signal PROPN cana-3719 292 16 processing processing NOUN cana-3719 292 17 , , PUNCT cana-3719 292 18 221 221 NUM cana-3719 292 19 , , PUNCT cana-3719 292 20 109504 109504 NUM cana-3719 292 21 . . PUNCT cana-3719 293 1 [ [ X cana-3719 293 2 23 23 NUM cana-3719 293 3 ] ] PUNCT cana-3719 293 4 . . PUNCT cana-3719 294 1 jia jia PROPN cana-3719 294 2 , , PUNCT cana-3719 294 3 z. z. PROPN cana-3719 294 4 , , PUNCT cana-3719 294 5 ng ng PROPN cana-3719 294 6 , , PUNCT cana-3719 294 7 m. m. PROPN cana-3719 294 8 k. k. PROPN cana-3719 294 9 , , PUNCT cana-3719 294 10 song song PROPN cana-3719 294 11 , , PUNCT cana-3719 294 12 g. g. PROPN cana-3719 294 13 j. j. PROPN cana-3719 294 14 ( ( PUNCT cana-3719 294 15 2019 2019 NUM cana-3719 294 16 ) ) PUNCT cana-3719 294 17 . . PUNCT cana-3719 295 1 robust robust ADJ cana-3719 295 2 quaternion quaternion NOUN cana-3719 295 3 matrix matrix NOUN cana-3719 295 4 completion completion NOUN cana-3719 295 5 with with ADP cana-3719 295 6 applications application NOUN cana-3719 295 7 to to ADP cana-3719 295 8 image image NOUN cana-3719 295 9 inpainting inpainte VERB cana-3719 295 10 . . PUNCT cana-3719 296 1 numerical numerical PROPN cana-3719 296 2 linear linear PROPN cana-3719 296 3 algebra algebra PROPN cana-3719 296 4 with with ADP cana-3719 296 5 applications application NOUN cana-3719 296 6 , , PUNCT cana-3719 296 7 26(4 26(4 NUM cana-3719 296 8 ) ) PUNCT cana-3719 296 9 , , PUNCT cana-3719 296 10 e2245 e2245 PROPN cana-3719 296 11 . . PUNCT cana-3719 297 1 [ [ X cana-3719 297 2 24 24 NUM cana-3719 297 3 ] ] PUNCT cana-3719 297 4 . . PUNCT cana-3719 298 1 zhu zhu PROPN cana-3719 298 2 , , PUNCT cana-3719 298 3 y. y. PROPN cana-3719 298 4 , , PUNCT cana-3719 298 5 wang wang PROPN cana-3719 298 6 , , PUNCT cana-3719 298 7 c. c. PROPN cana-3719 298 8 , , PUNCT cana-3719 298 9 geng geng PROPN cana-3719 298 10 , , PUNCT cana-3719 298 11 s. s. PROPN cana-3719 298 12 , , PUNCT cana-3719 298 13 yu yu PROPN cana-3719 298 14 , , PUNCT cana-3719 298 15 y. y. PROPN cana-3719 298 16 , , PUNCT cana-3719 298 17 hao hao PROPN cana-3719 298 18 , , PUNCT cana-3719 298 19 x. x. NOUN cana-3719 298 20 ( ( PUNCT cana-3719 298 21 2022 2022 NUM cana-3719 298 22 ) ) PUNCT cana-3719 298 23 . . PUNCT cana-3719 299 1 multi multi ADJ cana-3719 299 2 - - ADJ cana-3719 299 3 scale scale ADJ cana-3719 299 4 gradient gradient ADJ cana-3719 299 5 attention attention NOUN cana-3719 299 6 guidance guidance NOUN cana-3719 299 7 and and CCONJ cana-3719 299 8 adaptive adaptive ADJ cana-3719 299 9 style style NOUN cana-3719 299 10 fusion fusion NOUN cana-3719 299 11 for for ADP cana-3719 299 12 image image NOUN cana-3719 299 13 inpainting inpainte VERB cana-3719 299 14 . . PUNCT cana-3719 300 1 journal journal PROPN cana-3719 300 2 of of ADP cana-3719 300 3 visual visual ADJ cana-3719 300 4 communication communication NOUN cana-3719 300 5 and and CCONJ cana-3719 300 6 image image NOUN cana-3719 300 7 representation representation NOUN cana-3719 300 8 , , PUNCT cana-3719 300 9 89 89 NUM cana-3719 300 10 , , PUNCT cana-3719 300 11 103681 103681 NUM cana-3719 300 12 . . PUNCT cana-3719 301 1 [ [ X cana-3719 301 2 25 25 NUM cana-3719 301 3 ] ] PUNCT cana-3719 301 4 . . PUNCT cana-3719 302 1 woo woo PROPN cana-3719 302 2 , , PUNCT cana-3719 302 3 s. s. PROPN cana-3719 302 4 , , PUNCT cana-3719 302 5 ko ko PROPN cana-3719 302 6 , , PUNCT cana-3719 302 7 k. k. PROPN cana-3719 302 8 , , PUNCT cana-3719 302 9 kim kim PROPN cana-3719 302 10 , , PUNCT cana-3719 302 11 c. c. PROPN cana-3719 302 12 s. s. PROPN cana-3719 302 13 ( ( PUNCT cana-3719 302 14 2024 2024 NUM cana-3719 302 15 ) ) PUNCT cana-3719 302 16 . . PUNCT cana-3719 303 1 local local ADJ cana-3719 303 2 and and CCONJ cana-3719 303 3 global global ADJ cana-3719 303 4 mixture mixture NOUN cana-3719 303 5 network network NOUN cana-3719 303 6 for for ADP cana-3719 303 7 image image NOUN cana-3719 303 8 inpainting inpainte VERB cana-3719 303 9 . . PUNCT cana-3719 304 1 journal journal PROPN cana-3719 304 2 of of ADP cana-3719 304 3 visual visual ADJ cana-3719 304 4 communication communication NOUN cana-3719 304 5 and and CCONJ cana-3719 304 6 image image NOUN cana-3719 304 7 representation representation NOUN cana-3719 304 8 , , PUNCT cana-3719 304 9 104 104 NUM cana-3719 304 10 , , PUNCT cana-3719 304 11 104312 104312 NUM cana-3719 304 12 . . PUNCT cana-3719 305 1 [ [ X cana-3719 305 2 26 26 NUM cana-3719 305 3 ] ] PUNCT cana-3719 305 4 . . PUNCT cana-3719 306 1 c. c. PROPN cana-3719 306 2 szegedy szegedy PROPN cana-3719 306 3 , , PUNCT cana-3719 306 4 w. w. PROPN cana-3719 306 5 liu liu PROPN cana-3719 306 6 , , PUNCT cana-3719 306 7 y. y. PROPN cana-3719 306 8 jia jia PROPN cana-3719 306 9 , , PUNCT cana-3719 306 10 p. p. PROPN cana-3719 306 11 sermanet sermanet NOUN cana-3719 306 12 , , PUNCT cana-3719 306 13 s. s. PROPN cana-3719 306 14 reed reed PROPN cana-3719 306 15 , , PUNCT cana-3719 306 16 d. d. PROPN cana-3719 306 17 anguelov anguelov PROPN cana-3719 306 18 , , PUNCT cana-3719 306 19 d. d. PROPN cana-3719 306 20 erhan erhan PROPN cana-3719 306 21 , , PUNCT cana-3719 306 22 v. v. PROPN cana-3719 306 23 vanhoucke vanhoucke PROPN cana-3719 306 24 , , PUNCT cana-3719 306 25 and and CCONJ cana-3719 306 26 a. a. NOUN cana-3719 306 27 rabinovich rabinovich NOUN cana-3719 306 28 , , PUNCT cana-3719 306 29 β€˜ ' PUNCT cana-3719 306 30 β€˜ ' PUNCT cana-3719 306 31 going go VERB cana-3719 306 32 deeper deeply ADV cana-3719 306 33 with with ADP cana-3719 306 34 convolutions convolution NOUN cana-3719 306 35 , , PUNCT cana-3719 306 36 ’’ ’' PUNCT cana-3719 306 37 in in ADP cana-3719 306 38 proc proc NOUN cana-3719 306 39 . . PUNCT cana-3719 307 1 ieee ieee NOUN cana-3719 307 2 conf conf NOUN cana-3719 307 3 . . PUNCT cana-3719 308 1 comput comput NOUN cana-3719 308 2 . . PUNCT cana-3719 309 1 vis vis X cana-3719 309 2 . . X cana-3719 309 3 pattern pattern NOUN cana-3719 309 4 recognit recognit VERB cana-3719 309 5 . . PUNCT cana-3719 310 1 ( ( PUNCT cana-3719 310 2 cvpr cvpr NOUN cana-3719 310 3 ) ) PUNCT cana-3719 310 4 , , PUNCT cana-3719 310 5 jun jun PROPN cana-3719 310 6 . . PROPN cana-3719 310 7 2015 2015 NUM cana-3719 310 8 , , PUNCT cana-3719 310 9 pp pp ADJ cana-3719 310 10 . . NOUN cana-3719 310 11 1 1 NUM cana-3719 310 12 – – PUNCT cana-3719 310 13 9 9 NUM cana-3719 310 14 . . PUNCT cana-3719 311 1 [ [ X cana-3719 311 2 27 27 NUM cana-3719 311 3 ] ] PUNCT cana-3719 311 4 . . PUNCT cana-3719 312 1 c. c. PROPN cana-3719 312 2 szegedy szegedy PROPN cana-3719 312 3 , , PUNCT cana-3719 312 4 v. v. PROPN cana-3719 312 5 vanhoucke vanhoucke PROPN cana-3719 312 6 , , PUNCT cana-3719 312 7 s. s. PROPN cana-3719 312 8 ioffe ioffe PROPN cana-3719 312 9 , , PUNCT cana-3719 312 10 j. j. PROPN cana-3719 312 11 shlens shlens PROPN cana-3719 312 12 , , PUNCT cana-3719 312 13 and and CCONJ cana-3719 312 14 z. z. PROPN cana-3719 312 15 wojna wojna PROPN cana-3719 312 16 , , PUNCT cana-3719 312 17 β€˜ ' PUNCT cana-3719 312 18 β€˜ ' PUNCT cana-3719 312 19 rethinking rethink VERB cana-3719 312 20 the the DET cana-3719 312 21 inception inception ADJ cana-3719 312 22 architecture architecture NOUN cana-3719 312 23 for for ADP cana-3719 312 24 computer computer NOUN cana-3719 312 25 vision vision NOUN cana-3719 312 26 , , PUNCT cana-3719 312 27 ’’ ’' PUNCT cana-3719 312 28 in in ADP cana-3719 312 29 proc proc NOUN cana-3719 312 30 . . PUNCT cana-3719 313 1 ieee ieee NOUN cana-3719 313 2 conf conf NOUN cana-3719 313 3 . . PUNCT cana-3719 314 1 comput comput NOUN cana-3719 314 2 . . PUNCT cana-3719 315 1 vis vis X cana-3719 315 2 . . X cana-3719 315 3 pattern pattern NOUN cana-3719 315 4 recognit recognit VERB cana-3719 315 5 . . PUNCT cana-3719 316 1 ( ( PUNCT cana-3719 316 2 cvpr cvpr NOUN cana-3719 316 3 ) ) PUNCT cana-3719 316 4 , , PUNCT cana-3719 316 5 jun jun PROPN cana-3719 316 6 . . PROPN cana-3719 316 7 2016 2016 NUM cana-3719 316 8 , , PUNCT cana-3719 316 9 pp pp ADJ cana-3719 316 10 . . PUNCT cana-3719 317 1 2818–2826 2818–2826 NUM cana-3719 317 2 . . PUNCT cana-3719 318 1 [ [ X cana-3719 318 2 28 28 NUM cana-3719 318 3 ] ] PUNCT cana-3719 318 4 . . PUNCT cana-3719 319 1 c. c. PROPN cana-3719 319 2 szegedy szegedy PROPN cana-3719 319 3 , , PUNCT cana-3719 319 4 s. s. PROPN cana-3719 319 5 ioffe ioffe PROPN cana-3719 319 6 , , PUNCT cana-3719 319 7 v. v. PROPN cana-3719 319 8 vanhoucke vanhoucke PROPN cana-3719 319 9 , , PUNCT cana-3719 319 10 and and CCONJ cana-3719 319 11 a. a. NOUN cana-3719 319 12 alemi alemi PROPN cana-3719 319 13 , , PUNCT cana-3719 319 14 β€˜ ' PUNCT cana-3719 319 15 β€˜ ' PUNCT cana-3719 319 16 inception inception ADJ cana-3719 319 17 -v4 -v4 ADJ cana-3719 319 18 , , PUNCT cana-3719 319 19 inception inception NOUN cana-3719 319 20 - - PUNCT cana-3719 319 21 resnet resnet NOUN cana-3719 319 22 and and CCONJ cana-3719 319 23 the the DET cana-3719 319 24 impact impact NOUN cana-3719 319 25 of of ADP cana-3719 319 26 residual residual ADJ cana-3719 319 27 connections connection NOUN cana-3719 319 28 on on ADP cana-3719 319 29 learning learn VERB cana-3719 319 30 , , PUNCT cana-3719 319 31 ’’ ’' PUNCT cana-3719 319 32 in in ADP cana-3719 319 33 proc proc NOUN cana-3719 319 34 . . PUNCT cana-3719 320 1 aaai aaai PROPN cana-3719 320 2 conf conf PROPN cana-3719 320 3 . . PUNCT cana-3719 321 1 artif artif PROPN cana-3719 321 2 . . PUNCT cana-3719 322 1 intell intell PROPN cana-3719 322 2 . . PUNCT cana-3719 322 3 , , PUNCT cana-3719 322 4 2017 2017 NUM cana-3719 322 5 , , PUNCT cana-3719 322 6 vol vol NOUN cana-3719 322 7 . . PROPN cana-3719 323 1 31 31 NUM cana-3719 323 2 , , PUNCT cana-3719 323 3 no no INTJ cana-3719 323 4 . . NOUN cana-3719 323 5 1 1 NUM cana-3719 323 6 , , PUNCT cana-3719 323 7 pp pp ADJ cana-3719 323 8 . . PUNCT cana-3719 324 1 1–7.b 1–7.b NUM cana-3719 325 1 [ [ X cana-3719 325 2 29 29 NUM cana-3719 325 3 ] ] PUNCT cana-3719 325 4 . . PUNCT cana-3719 326 1 o. o. PROPN cana-3719 326 2 russakovsky russakovsky PROPN cana-3719 326 3 , , PUNCT cana-3719 326 4 j. j. PROPN cana-3719 326 5 deng deng PROPN cana-3719 326 6 , , PUNCT cana-3719 326 7 h. h. PROPN cana-3719 326 8 su su PROPN cana-3719 326 9 , , PUNCT cana-3719 326 10 j. j. PROPN cana-3719 326 11 krause krause PROPN cana-3719 326 12 , , PUNCT cana-3719 326 13 s. s. PROPN cana-3719 326 14 satheesh satheesh PROPN cana-3719 326 15 , , PUNCT cana-3719 326 16 s. s. PROPN cana-3719 326 17 ma ma PROPN cana-3719 326 18 , , PUNCT cana-3719 326 19 z. z. PROPN cana-3719 326 20 huang huang PROPN cana-3719 326 21 , , PUNCT cana-3719 326 22 a. a. PROPN cana-3719 326 23 karpathy karpathy PROPN cana-3719 326 24 , , PUNCT cana-3719 326 25 a. a. NOUN cana-3719 326 26 khosla khosla PROPN cana-3719 326 27 , , PUNCT cana-3719 326 28 m. m. PROPN cana-3719 326 29 bernstein bernstein PROPN cana-3719 326 30 , , PUNCT cana-3719 326 31 a. a. PROPN cana-3719 326 32 c. c. PROPN cana-3719 326 33 berg berg PROPN cana-3719 326 34 , , PUNCT cana-3719 326 35 and and CCONJ cana-3719 326 36 l. l. PROPN cana-3719 326 37 fei fei PROPN cana-3719 326 38 - - PUNCT cana-3719 326 39 fei fei PROPN cana-3719 326 40 , , PUNCT cana-3719 326 41 β€˜ ' PUNCT cana-3719 326 42 β€˜ ' PUNCT cana-3719 326 43 imagenet imagenet ADJ cana-3719 326 44 large large ADJ cana-3719 326 45 scale scale NOUN cana-3719 326 46 visual visual ADJ cana-3719 326 47 recognition recognition NOUN cana-3719 326 48 challenge challenge NOUN cana-3719 326 49 , , PUNCT cana-3719 326 50 ’’ ’' PUNCT cana-3719 326 51 int int NOUN cana-3719 326 52 . . PUNCT cana-3719 327 1 j. j. PROPN cana-3719 327 2 comput comput PROPN cana-3719 327 3 . . PUNCT cana-3719 328 1 vis vis X cana-3719 328 2 . . PROPN cana-3719 328 3 , , PUNCT cana-3719 328 4 vol vol NOUN cana-3719 328 5 . . PROPN cana-3719 328 6 115 115 NUM cana-3719 328 7 , , PUNCT cana-3719 328 8 no no INTJ cana-3719 328 9 . . NOUN cana-3719 328 10 3 3 NUM cana-3719 328 11 , , PUNCT cana-3719 328 12 pp pp ADJ cana-3719 328 13 . . PUNCT cana-3719 329 1 211–252 211–252 NUM cana-3719 329 2 , , PUNCT cana-3719 329 3 dec dec PROPN cana-3719 329 4 . . PROPN cana-3719 329 5 2015 2015 NUM cana-3719 329 6 . . PUNCT cana-3719 330 1 [ [ X cana-3719 330 2 30 30 NUM cana-3719 330 3 ] ] PUNCT cana-3719 330 4 . . PUNCT cana-3719 331 1 k. k. PROPN cana-3719 331 2 simonyan simonyan PROPN cana-3719 331 3 and and CCONJ cana-3719 331 4 a. a. NOUN cana-3719 331 5 zisserman zisserman PROPN cana-3719 331 6 , , PUNCT cana-3719 331 7 β€˜ ' PUNCT cana-3719 331 8 β€˜ ' PUNCT cana-3719 331 9 very very ADV cana-3719 331 10 deep deep ADJ cana-3719 331 11 convolutional convolutional ADJ cana-3719 331 12 networks network NOUN cana-3719 331 13 for for ADP cana-3719 331 14 large large ADJ cana-3719 331 15 -scale -scale ADJ cana-3719 331 16 image image NOUN cana-3719 331 17 recognition recognition NOUN cana-3719 331 18 , , PUNCT cana-3719 331 19 ’’ ’' PUNCT cana-3719 331 20 2014 2014 NUM cana-3719 331 21 , , PUNCT cana-3719 331 22 arxiv:1409.1556 arxiv:1409.1556 NOUN cana-3719 331 23 . . PUNCT cana-3719 332 1 [ [ X cana-3719 332 2 31 31 NUM cana-3719 332 3 ] ] PUNCT cana-3719 332 4 . . PUNCT cana-3719 333 1 wang wang PROPN cana-3719 333 2 , , PUNCT cana-3719 333 3 w. w. PROPN cana-3719 333 4 , , PUNCT cana-3719 333 5 niu niu PROPN cana-3719 333 6 , , PUNCT cana-3719 333 7 l. l. PROPN cana-3719 333 8 , , PUNCT cana-3719 333 9 zhang zhang PROPN cana-3719 333 10 , , PUNCT cana-3719 333 11 j. j. PROPN cana-3719 333 12 , , PUNCT cana-3719 333 13 yang yang PROPN cana-3719 333 14 , , PUNCT cana-3719 333 15 x. x. PROPN cana-3719 333 16 , , PUNCT cana-3719 333 17 zhang zhang PROPN cana-3719 333 18 , , PUNCT cana-3719 333 19 l. l. PROPN cana-3719 333 20 ( ( PUNCT cana-3719 333 21 2022 2022 NUM cana-3719 333 22 ) ) PUNCT cana-3719 333 23 . . PUNCT cana-3719 334 1 dual dual ADJ cana-3719 334 2 -path -path NOUN cana-3719 334 3 image image NOUN cana-3719 334 4 inpainting inpainte VERB cana-3719 334 5 with with ADP cana-3719 334 6 auxiliary auxiliary ADJ cana-3719 334 7 gan gan ADJ cana-3719 334 8 inversion inversion NOUN cana-3719 334 9 . . PUNCT cana-3719 335 1 in in ADP cana-3719 335 2 proceedings proceeding NOUN cana-3719 335 3 of of ADP cana-3719 335 4 the the DET cana-3719 335 5 ieee ieee NOUN cana-3719 335 6 / / SYM cana-3719 335 7 cvf cvf NOUN cana-3719 335 8 conference conference NOUN cana-3719 335 9 on on ADP cana-3719 335 10 computer computer NOUN cana-3719 335 11 vision vision NOUN cana-3719 335 12 and and CCONJ cana-3719 335 13 pattern pattern NOUN cana-3719 335 14 recognition recognition NOUN cana-3719 335 15 ( ( PUNCT cana-3719 335 16 pp pp ADJ cana-3719 335 17 . . PUNCT cana-3719 335 18 1142111430 1142111430 NUM cana-3719 335 19 ) ) PUNCT cana-3719 335 20 . . PUNCT cana-3719 336 1 [ [ X cana-3719 336 2 32 32 NUM cana-3719 336 3 ] ] PUNCT cana-3719 336 4 . . PUNCT cana-3719 337 1 li li PROPN cana-3719 337 2 , , PUNCT cana-3719 337 3 z. z. PROPN cana-3719 337 4 , , PUNCT cana-3719 337 5 zhang zhang PROPN cana-3719 337 6 , , PUNCT cana-3719 337 7 y. y. PROPN cana-3719 337 8 , , PUNCT cana-3719 337 9 du du PROPN cana-3719 337 10 , , PUNCT cana-3719 337 11 y. y. PROPN cana-3719 337 12 , , PUNCT cana-3719 337 13 wang wang PROPN cana-3719 337 14 , , PUNCT cana-3719 337 15 x. x. PROPN cana-3719 337 16 , , PUNCT cana-3719 337 17 wen wen PROPN cana-3719 337 18 , , PUNCT cana-3719 337 19 c. c. PROPN cana-3719 337 20 , , PUNCT cana-3719 337 21 zhang zhang PROPN cana-3719 337 22 , , PUNCT cana-3719 337 23 y. y. PROPN cana-3719 337 24 , , PUNCT cana-3719 337 25 ... ... PUNCT cana-3719 338 1 jia jia PROPN cana-3719 338 2 , , PUNCT cana-3719 338 3 f. f. PROPN cana-3719 338 4 ( ( PUNCT cana-3719 338 5 2024 2024 NUM cana-3719 338 6 ) ) PUNCT cana-3719 338 7 . . PUNCT cana-3719 339 1 stnet stnet NOUN cana-3719 339 2 : : PUNCT cana-3719 339 3 structure structure NOUN cana-3719 339 4 and and CCONJ cana-3719 339 5 texture texture NOUN cana-3719 339 6 -guided -guided ADJ cana-3719 339 7 network network NOUN cana-3719 339 8 for for ADP cana-3719 339 9 image image NOUN cana-3719 339 10 inpainting inpainte VERB cana-3719 339 11 . . PUNCT cana-3719 340 1 pattern pattern NOUN cana-3719 340 2 recognition recognition NOUN cana-3719 340 3 , , PUNCT cana-3719 340 4 156 156 NUM cana-3719 340 5 , , PUNCT cana-3719 340 6 110786 110786 NUM cana-3719 340 7 . . PUNCT cana-3719 341 1 [ [ X cana-3719 341 2 33 33 NUM cana-3719 341 3 ] ] PUNCT cana-3719 341 4 . . PUNCT cana-3719 342 1 rakibe rakibe PROPN cana-3719 342 2 , , PUNCT cana-3719 342 3 p. p. PROPN cana-3719 342 4 l. l. PROPN cana-3719 342 5 , , PUNCT cana-3719 342 6 patil patil PROPN cana-3719 342 7 , , PUNCT cana-3719 342 8 p. p. PROPN cana-3719 342 9 d. d. PROPN cana-3719 342 10 ( ( PUNCT cana-3719 342 11 2024 2024 NUM cana-3719 342 12 ) ) PUNCT cana-3719 342 13 . . PUNCT cana-3719 343 1 improved improve VERB cana-3719 343 2 medical medical ADJ cana-3719 343 3 image image NOUN cana-3719 343 4 inpainting inpainte VERB cana-3719 343 5 using use VERB cana-3719 343 6 automatic automatic ADJ cana-3719 343 7 multi multi ADJ cana-3719 343 8 -task -task PROPN cana-3719 343 9 learning learn VERB cana-3719 343 10 driven drive VERB cana-3719 343 11 deep deep ADJ cana-3719 343 12 learning learning NOUN cana-3719 343 13 approach approach NOUN cana-3719 343 14 . . PUNCT cana-3719 344 1 e e X cana-3719 344 2 - - ADJ cana-3719 344 3 prime prime ADJ cana-3719 344 4 - - PUNCT cana-3719 344 5 advances advance NOUN cana-3719 344 6 in in ADP cana-3719 344 7 electrical electrical ADJ cana-3719 344 8 engineering engineering NOUN cana-3719 344 9 , , PUNCT cana-3719 344 10 electronics electronic NOUN cana-3719 344 11 and and CCONJ cana-3719 344 12 energy energy NOUN cana-3719 344 13 , , PUNCT cana-3719 344 14 9 9 NUM cana-3719 344 15 , , PUNCT cana-3719 344 16 100678 100678 NUM cana-3719 344 17 . . PUNCT cana-3719 345 1 [ [ X cana-3719 345 2 34 34 NUM cana-3719 345 3 ] ] PUNCT cana-3719 345 4 . . PUNCT cana-3719 346 1 yeh yeh PROPN cana-3719 346 2 , , PUNCT cana-3719 346 3 c. c. PROPN cana-3719 346 4 h. h. PROPN cana-3719 346 5 , , PUNCT cana-3719 346 6 yang yang PROPN cana-3719 346 7 , , PUNCT cana-3719 346 8 h. h. PROPN cana-3719 346 9 f. f. PROPN cana-3719 346 10 , , PUNCT cana-3719 346 11 chen chen PROPN cana-3719 346 12 , , PUNCT cana-3719 346 13 m. m. PROPN cana-3719 346 14 j. j. PROPN cana-3719 346 15 , , PUNCT cana-3719 346 16 kang kang PROPN cana-3719 346 17 , , PUNCT cana-3719 346 18 l. l. PROPN cana-3719 346 19 w. w. PROPN cana-3719 346 20 ( ( PUNCT cana-3719 346 21 2024 2024 NUM cana-3719 346 22 ) ) PUNCT cana-3719 346 23 . . PUNCT cana-3719 347 1 image image NOUN cana-3719 347 2 inpainting inpainte VERB cana-3719 347 3 based base VERB cana-3719 347 4 on on ADP cana-3719 347 5 gan gan NOUN cana-3719 347 6 - - PUNCT cana-3719 347 7 driven drive VERB cana-3719 347 8 structure structure NOUN cana-3719 347 9 - - PUNCT cana-3719 347 10 and and CCONJ cana-3719 347 11 texture texture ADJ cana-3719 347 12 - - PUNCT cana-3719 347 13 aware aware ADJ cana-3719 347 14 learning learning NOUN cana-3719 347 15 with with ADP cana-3719 347 16 application application NOUN cana-3719 347 17 to to PART cana-3719 347 18 object object VERB cana-3719 347 19 removal removal NOUN cana-3719 347 20 . . PUNCT cana-3719 347 21 applied apply VERB cana-3719 347 22 soft soft ADJ cana-3719 347 23 computing computing NOUN cana-3719 347 24 , , PUNCT cana-3719 347 25 161 161 NUM cana-3719 347 26 , , PUNCT cana-3719 347 27 111748 111748 NUM cana-3719 347 28 . . PUNCT cana-3719 348 1 [ [ X cana-3719 348 2 35 35 NUM cana-3719 348 3 ] ] PUNCT cana-3719 348 4 . . PUNCT cana-3719 348 5 x. x. PROPN cana-3719 348 6 mao mao PROPN cana-3719 348 7 , , PUNCT cana-3719 348 8 q. q. PROPN cana-3719 348 9 li li PROPN cana-3719 348 10 , , PUNCT cana-3719 348 11 h. h. PROPN cana-3719 348 12 xie xie PROPN cana-3719 348 13 , , PUNCT cana-3719 348 14 r. r. PROPN cana-3719 348 15 y. y. PROPN cana-3719 348 16 k. k. PROPN cana-3719 348 17 lau lau PROPN cana-3719 348 18 , , PUNCT cana-3719 348 19 z. z. PROPN cana-3719 348 20 wang wang PROPN cana-3719 348 21 , , PUNCT cana-3719 348 22 and and CCONJ cana-3719 348 23 s. s. PROPN cana-3719 348 24 p. p. PROPN cana-3719 348 25 smolley smolley PROPN cana-3719 348 26 , , PUNCT cana-3719 348 27 β€˜ ' PUNCT cana-3719 348 28 β€˜ ' PUNCT cana-3719 348 29 least least ADJ cana-3719 348 30 squares square NOUN cana-3719 348 31 generative generative VERB cana-3719 348 32 adversarial adversarial ADJ cana-3719 348 33 networks network NOUN cana-3719 348 34 , , PUNCT cana-3719 348 35 ’’ ’' PUNCT cana-3719 348 36 in in ADP cana-3719 348 37 proc proc NOUN cana-3719 348 38 . . PUNCT cana-3719 349 1 ieee ieee PROPN cana-3719 349 2 int int PROPN cana-3719 349 3 . . PUNCT cana-3719 350 1 conf conf PROPN cana-3719 350 2 . . PUNCT cana-3719 351 1 comput comput NOUN cana-3719 351 2 . . PUNCT cana-3719 352 1 vis vis X cana-3719 352 2 . . PUNCT cana-3719 352 3 ( ( PUNCT cana-3719 352 4 iccv iccv PROPN cana-3719 352 5 ) ) PUNCT cana-3719 352 6 , , PUNCT cana-3719 352 7 oct oct PROPN cana-3719 352 8 . . PROPN cana-3719 352 9 2017 2017 NUM cana-3719 352 10 , , PUNCT cana-3719 352 11 pp pp ADJ cana-3719 352 12 . . PUNCT cana-3719 352 13 2813–2821 2813–2821 NUM cana-3719 352 14 . . PUNCT cana-3719 353 1 [ [ X cana-3719 353 2 36 36 NUM cana-3719 353 3 ] ] PUNCT cana-3719 353 4 . . PUNCT cana-3719 354 1 i. i. PROPN cana-3719 354 2 gulrajani gulrajani PROPN cana-3719 354 3 , , PUNCT cana-3719 354 4 f. f. PROPN cana-3719 354 5 ahmed ahmed PROPN cana-3719 354 6 , , PUNCT cana-3719 354 7 m. m. NOUN cana-3719 354 8 arjovsky arjovsky PROPN cana-3719 354 9 , , PUNCT cana-3719 354 10 v. v. CCONJ cana-3719 354 11 dumoulin dumoulin PROPN cana-3719 354 12 , , PUNCT cana-3719 354 13 and and CCONJ cana-3719 354 14 a. a. PROPN cana-3719 354 15 c. c. PROPN cana-3719 354 16 courville courville PROPN cana-3719 354 17 , , PUNCT cana-3719 354 18 β€˜ ' PUNCT cana-3719 354 19 β€˜ ' PUNCT cana-3719 354 20 improved improved ADJ cana-3719 354 21 training training NOUN cana-3719 354 22 of of ADP cana-3719 354 23 wasserstein wasserstein NOUN cana-3719 354 24 gans gan NOUN cana-3719 354 25 , , PUNCT cana-3719 354 26 ’’ ’' PUNCT cana-3719 354 27 in in ADP cana-3719 354 28 proc proc NOUN cana-3719 354 29 . . PUNCT cana-3719 355 1 adv adv PROPN cana-3719 355 2 . . PUNCT cana-3719 355 3 neural neural PROPN cana-3719 355 4 inf inf PROPN cana-3719 355 5 . . PUNCT cana-3719 355 6 process process NOUN cana-3719 355 7 . . PUNCT cana-3719 356 1 syst syst PROPN cana-3719 356 2 . . PUNCT cana-3719 356 3 , , PUNCT cana-3719 356 4 vol vol NOUN cana-3719 356 5 . . PROPN cana-3719 356 6 30 30 NUM cana-3719 356 7 , , PUNCT cana-3719 356 8 2017 2017 NUM cana-3719 356 9 , , PUNCT cana-3719 356 10 pp pp ADJ cana-3719 356 11 . . PUNCT cana-3719 357 1 1–11 1–11 NOUN cana-3719 357 2 . . PUNCT cana-3719 358 1 [ [ X cana-3719 358 2 37 37 NUM cana-3719 358 3 ] ] PUNCT cana-3719 358 4 . . PUNCT cana-3719 358 5 m. m. PROPN cana-3719 358 6 arjovsky arjovsky PROPN cana-3719 358 7 , , PUNCT cana-3719 358 8 s. s. PROPN cana-3719 358 9 chintala chintala PROPN cana-3719 358 10 , , PUNCT cana-3719 358 11 and and CCONJ cana-3719 358 12 l. l. PROPN cana-3719 358 13 bottou bottou PROPN cana-3719 358 14 , , PUNCT cana-3719 358 15 β€˜ ' PUNCT cana-3719 358 16 β€˜ ' PUNCT cana-3719 358 17 wasserstein wasserstein NOUN cana-3719 358 18 generative generative VERB cana-3719 358 19 adversarial adversarial ADJ cana-3719 358 20 networks network NOUN cana-3719 358 21 , , PUNCT cana-3719 358 22 ’’ ’' PUNCT cana-3719 358 23 in in ADP cana-3719 358 24 proc proc NOUN cana-3719 358 25 . . PUNCT cana-3719 359 1 int int NOUN cana-3719 359 2 . . PUNCT cana-3719 359 3 conf conf PROPN cana-3719 359 4 . . PUNCT cana-3719 360 1 mach mach PROPN cana-3719 360 2 . . PUNCT cana-3719 361 1 learn learn VERB cana-3719 361 2 . . PROPN cana-3719 361 3 , , PUNCT cana-3719 361 4 2017 2017 NUM cana-3719 361 5 , , PUNCT cana-3719 361 6 pp pp ADJ cana-3719 361 7 . . PUNCT cana-3719 362 1 214–223 214–223 NUM cana-3719 362 2 . . PUNCT cana-3719 363 1 [ [ X cana-3719 363 2 38 38 NUM cana-3719 363 3 ] ] PUNCT cana-3719 363 4 . . PUNCT cana-3719 364 1 huang huang PROPN cana-3719 364 2 , , PUNCT cana-3719 364 3 w. w. PROPN cana-3719 364 4 , , PUNCT cana-3719 364 5 deng deng PROPN cana-3719 364 6 , , PUNCT cana-3719 364 7 y. y. PROPN cana-3719 364 8 , , PUNCT cana-3719 364 9 hui hui PROPN cana-3719 364 10 , , PUNCT cana-3719 364 11 s. s. PROPN cana-3719 364 12 , , PUNCT cana-3719 364 13 wu wu PROPN cana-3719 364 14 , , PUNCT cana-3719 364 15 y. y. PROPN cana-3719 364 16 , , PUNCT cana-3719 364 17 zhou zhou PROPN cana-3719 364 18 , , PUNCT cana-3719 364 19 s. s. PROPN cana-3719 364 20 , , PUNCT cana-3719 364 21 wang wang PROPN cana-3719 364 22 , , PUNCT cana-3719 364 23 j. j. PROPN cana-3719 364 24 ( ( PUNCT cana-3719 364 25 2024 2024 NUM cana-3719 364 26 ) ) PUNCT cana-3719 364 27 . . PUNCT cana-3719 365 1 sparse sparse VERB cana-3719 365 2 self self NOUN cana-3719 365 3 - - PUNCT cana-3719 365 4 attention attention NOUN cana-3719 365 5 transformer transformer NOUN cana-3719 365 6 for for ADP cana-3719 365 7 image image NOUN cana-3719 365 8 inpainting inpainte VERB cana-3719 365 9 . . PUNCT cana-3719 366 1 pattern pattern NOUN cana-3719 366 2 recognition recognition NOUN cana-3719 366 3 , , PUNCT cana-3719 366 4 145 145 NUM cana-3719 366 5 , , PUNCT cana-3719 366 6 109897 109897 NUM cana-3719 366 7 . . PUNCT cana-3719 367 1 communications communication NOUN cana-3719 367 2 on on ADP cana-3719 367 3 applied apply VERB cana-3719 367 4 nonlinear nonlinear ADJ cana-3719 367 5 analysis analysis NOUN cana-3719 367 6 issn issn NOUN cana-3719 367 7 : : PUNCT cana-3719 367 8 1074 1074 NUM cana-3719 367 9 - - PUNCT cana-3719 367 10 133x 133x NUM cana-3719 367 11 vol vol NOUN cana-3719 367 12 32 32 NUM cana-3719 367 13 no no NOUN cana-3719 367 14 . . PUNCT cana-3719 368 1 8s 8s PROPN cana-3719 368 2 ( ( PUNCT cana-3719 368 3 2025 2025 NUM cana-3719 368 4 ) ) PUNCT cana-3719 368 5 586 586 NUM cana-3719 368 6 https://internationalpubls.com https://internationalpubls.com X cana-3719 369 1 [ [ X cana-3719 369 2 39 39 NUM cana-3719 369 3 ] ] PUNCT cana-3719 369 4 . . PUNCT cana-3719 369 5 q. q. PROPN cana-3719 369 6 dong dong PROPN cana-3719 369 7 , , PUNCT cana-3719 369 8 c. c. PROPN cana-3719 369 9 cao cao PROPN cana-3719 369 10 , , PUNCT cana-3719 369 11 and and CCONJ cana-3719 369 12 y. y. PROPN cana-3719 369 13 fu fu PROPN cana-3719 369 14 , , PUNCT cana-3719 369 15 β€˜ ' PUNCT cana-3719 369 16 β€˜ ' PUNCT cana-3719 369 17 incremental incremental ADJ cana-3719 369 18 transformer transformer NOUN cana-3719 369 19 structure structure NOUN cana-3719 369 20 enhanced enhance VERB cana-3719 369 21 image image NOUN cana-3719 369 22 inpainting inpainte VERB cana-3719 369 23 with with ADP cana-3719 369 24 masking mask VERB cana-3719 369 25 positional positional ADJ cana-3719 369 26 encoding encoding NOUN cana-3719 369 27 , , PUNCT cana-3719 369 28 ’’ ’' PUNCT cana-3719 369 29 in in ADP cana-3719 369 30 proc proc NOUN cana-3719 369 31 . . PUNCT cana-3719 370 1 ieee ieee NOUN cana-3719 370 2 / / SYM cana-3719 370 3 cvf cvf NOUN cana-3719 370 4 conf conf NOUN cana-3719 370 5 . . PUNCT cana-3719 371 1 comput comput NOUN cana-3719 371 2 . . PUNCT cana-3719 372 1 vis vis X cana-3719 372 2 . . X cana-3719 372 3 pattern pattern NOUN cana-3719 372 4 recognit recognit VERB cana-3719 372 5 . . PUNCT cana-3719 373 1 ( ( PUNCT cana-3719 373 2 cvpr cvpr NOUN cana-3719 373 3 ) ) PUNCT cana-3719 373 4 , , PUNCT cana-3719 373 5 jun jun PROPN cana-3719 373 6 . . PROPN cana-3719 373 7 2022 2022 NUM cana-3719 373 8 , , PUNCT cana-3719 373 9 pp pp ADJ cana-3719 373 10 . . PUNCT cana-3719 374 1 11348–11358 11348–11358 NOUN cana-3719 374 2 . . PUNCT cana-3719 375 1 [ [ X cana-3719 375 2 40 40 NUM cana-3719 375 3 ] ] PUNCT cana-3719 375 4 . . PUNCT cana-3719 376 1 k. k. PROPN cana-3719 377 1 he he PRON cana-3719 377 2 , , PUNCT cana-3719 377 3 x. x. PROPN cana-3719 377 4 chen chen PROPN cana-3719 377 5 , , PUNCT cana-3719 377 6 s. s. PROPN cana-3719 377 7 xie xie PROPN cana-3719 377 8 , , PUNCT cana-3719 377 9 y. y. PROPN cana-3719 377 10 li li PROPN cana-3719 377 11 , , PUNCT cana-3719 377 12 p. p. NOUN cana-3719 377 13 dollΓ‘r dollΓ‘r NOUN cana-3719 377 14 , , PUNCT cana-3719 377 15 and and CCONJ cana-3719 377 16 r. r. PROPN cana-3719 377 17 girshick girshick PROPN cana-3719 377 18 , , PUNCT cana-3719 377 19 β€˜ ' PUNCT cana-3719 377 20 β€˜ ' PUNCT cana-3719 377 21 masked masked ADJ cana-3719 377 22 autoencoders autoencoder NOUN cana-3719 377 23 are be AUX cana-3719 377 24 scalable scalable ADJ cana-3719 377 25 vision vision NOUN cana-3719 377 26 learners learner NOUN cana-3719 377 27 , , PUNCT cana-3719 377 28 ’’ ’' PUNCT cana-3719 377 29 in in ADP cana-3719 377 30 proc proc NOUN cana-3719 377 31 . . PUNCT cana-3719 378 1 ieee ieee NOUN cana-3719 378 2 / / SYM cana-3719 378 3 cvf cvf NOUN cana-3719 378 4 conf conf NOUN cana-3719 378 5 . . PUNCT cana-3719 379 1 comput comput NOUN cana-3719 379 2 . . PUNCT cana-3719 380 1 vis vis X cana-3719 380 2 . . X cana-3719 380 3 pattern pattern NOUN cana-3719 380 4 recognit recognit VERB cana-3719 380 5 . . PUNCT cana-3719 381 1 ( ( PUNCT cana-3719 381 2 cvpr cvpr NOUN cana-3719 381 3 ) ) PUNCT cana-3719 381 4 , , PUNCT cana-3719 381 5 jun jun PROPN cana-3719 381 6 . . PROPN cana-3719 381 7 2022 2022 NUM cana-3719 381 8 , , PUNCT cana-3719 381 9 pp pp ADV cana-3719 381 10 . . PUNCT cana-3719 382 1 15979–15988 15979–15988 NUM cana-3719 382 2 . . PUNCT cana-3719 383 1 [ [ X cana-3719 383 2 41 41 NUM cana-3719 383 3 ] ] PUNCT cana-3719 383 4 . . PUNCT cana-3719 384 1 y. y. PROPN cana-3719 384 2 zhou zhou PROPN cana-3719 384 3 , , PUNCT cana-3719 384 4 c. c. PROPN cana-3719 384 5 barnes barnes PROPN cana-3719 384 6 , , PUNCT cana-3719 384 7 e. e. PROPN cana-3719 384 8 shechtman shechtman PROPN cana-3719 384 9 , , PUNCT cana-3719 384 10 and and CCONJ cana-3719 384 11 s. s. PROPN cana-3719 384 12 amirghodsi amirghodsi PROPN cana-3719 384 13 , , PUNCT cana-3719 384 14 β€˜ ' PUNCT cana-3719 384 15 β€˜ ' PUNCT cana-3719 384 16 transfill transfill ADJ cana-3719 384 17 : : PUNCT cana-3719 384 18 reference reference NOUN cana-3719 384 19 -guided -guided ADJ cana-3719 384 20 image image NOUN cana-3719 384 21 inpainting inpainte VERB cana-3719 384 22 by by ADP cana-3719 384 23 merging merge VERB cana-3719 384 24 multiple multiple ADJ cana-3719 384 25 color color NOUN cana-3719 384 26 and and CCONJ cana-3719 384 27 spatial spatial ADJ cana-3719 384 28 transformations transformation NOUN cana-3719 384 29 , , PUNCT cana-3719 384 30 ’’ ’' PUNCT cana-3719 384 31 in in ADP cana-3719 384 32 proc proc NOUN cana-3719 384 33 . . PUNCT cana-3719 384 34 ieee ieee NOUN cana-3719 384 35 / / SYM cana-3719 384 36 cvf cvf NOUN cana-3719 384 37 conf conf NOUN cana-3719 384 38 . . PUNCT cana-3719 385 1 comput comput NOUN cana-3719 385 2 . . PUNCT cana-3719 386 1 vis vis X cana-3719 386 2 . . X cana-3719 386 3 pattern pattern NOUN cana-3719 386 4 recognit recognit VERB cana-3719 386 5 . . PUNCT cana-3719 387 1 ( ( PUNCT cana-3719 387 2 cvpr cvpr NOUN cana-3719 387 3 ) ) PUNCT cana-3719 387 4 , , PUNCT cana-3719 387 5 jun jun PROPN cana-3719 387 6 . . PROPN cana-3719 387 7 2021 2021 NUM cana-3719 387 8 , , PUNCT cana-3719 387 9 pp pp ADJ cana-3719 387 10 . . PUNCT cana-3719 387 11 2266–2267 2266–2267 NUM cana-3719 387 12 . . PUNCT cana-3719 388 1 [ [ X cana-3719 388 2 42 42 NUM cana-3719 388 3 ] ] PUNCT cana-3719 388 4 . . PUNCT cana-3719 389 1 campana campana PROPN cana-3719 389 2 , , PUNCT cana-3719 389 3 j. j. PROPN cana-3719 389 4 l. l. PROPN cana-3719 389 5 f. f. PROPN cana-3719 389 6 , , PUNCT cana-3719 389 7 decker decker PROPN cana-3719 389 8 , , PUNCT cana-3719 389 9 l. l. PROPN cana-3719 389 10 g. g. PROPN cana-3719 389 11 l. l. PROPN cana-3719 389 12 , , PUNCT cana-3719 389 13 e e PROPN cana-3719 389 14 souza souza PROPN cana-3719 389 15 , , PUNCT cana-3719 389 16 m. m. PROPN cana-3719 389 17 r. r. PROPN cana-3719 389 18 , , PUNCT cana-3719 389 19 de de PROPN cana-3719 389 20 almeida almeida PROPN cana-3719 389 21 maia maia PROPN cana-3719 389 22 , , PUNCT cana-3719 389 23 h. h. PROPN cana-3719 389 24 , , PUNCT cana-3719 389 25 pedrini pedrini PROPN cana-3719 389 26 , , PUNCT cana-3719 389 27 h. h. PROPN cana-3719 389 28 ( ( PUNCT cana-3719 389 29 2023 2023 NUM cana-3719 389 30 ) ) PUNCT cana-3719 389 31 . . PUNCT cana-3719 390 1 variable variable ADJ cana-3719 390 2 hyperparameter hyperparameter PROPN cana-3719 390 3 visual visual ADJ cana-3719 390 4 transformer transformer NOUN cana-3719 390 5 for for ADP cana-3719 390 6 efficient efficient ADJ cana-3719 390 7 image image NOUN cana-3719 390 8 inpainting inpainte VERB cana-3719 390 9 . . PUNCT cana-3719 391 1 computers computer NOUN cana-3719 391 2 & & CCONJ cana-3719 391 3 graphics graphic NOUN cana-3719 391 4 , , PUNCT cana-3719 391 5 113 113 NUM cana-3719 391 6 , , PUNCT cana-3719 391 7 57 57 NUM cana-3719 391 8 - - SYM cana-3719 391 9 68 68 NUM cana-3719 391 10 . . PUNCT cana-3719 392 1 [ [ X cana-3719 392 2 43 43 NUM cana-3719 392 3 ] ] PUNCT cana-3719 392 4 . . PUNCT cana-3719 393 1 chen chen PROPN cana-3719 393 2 , , PUNCT cana-3719 393 3 y. y. PROPN cana-3719 393 4 , , PUNCT cana-3719 393 5 zhang zhang PROPN cana-3719 393 6 , , PUNCT cana-3719 393 7 h. h. PROPN cana-3719 393 8 , , PUNCT cana-3719 393 9 liu liu PROPN cana-3719 393 10 , , PUNCT cana-3719 393 11 l. l. PROPN cana-3719 393 12 , , PUNCT cana-3719 393 13 chen chen PROPN cana-3719 393 14 , , PUNCT cana-3719 393 15 x. x. PROPN cana-3719 393 16 , , PUNCT cana-3719 393 17 zhang zhang PROPN cana-3719 393 18 , , PUNCT cana-3719 393 19 q. q. PROPN cana-3719 393 20 , , PUNCT cana-3719 393 21 yang yang PROPN cana-3719 393 22 , , PUNCT cana-3719 393 23 k. k. PROPN cana-3719 393 24 , , PUNCT cana-3719 393 25 ... ... PUNCT cana-3719 394 1 xie xie PROPN cana-3719 394 2 , , PUNCT cana-3719 394 3 j. j. PROPN cana-3719 394 4 ( ( PUNCT cana-3719 394 5 2021 2021 NUM cana-3719 394 6 ) ) PUNCT cana-3719 394 7 . . PUNCT cana-3719 395 1 research research NOUN cana-3719 395 2 on on ADP cana-3719 395 3 image image NOUN cana-3719 395 4 inpainting inpainte VERB cana-3719 395 5 algorithm algorithm NOUN cana-3719 395 6 of of ADP cana-3719 395 7 improved improved ADJ cana-3719 395 8 gan gan NOUN cana-3719 395 9 based base VERB cana-3719 395 10 on on ADP cana-3719 395 11 two two NUM cana-3719 395 12 - - PUNCT cana-3719 395 13 discriminations discrimination NOUN cana-3719 395 14 networks network NOUN cana-3719 395 15 . . PUNCT cana-3719 396 1 applied apply VERB cana-3719 396 2 intelligence intelligence NOUN cana-3719 396 3 , , PUNCT cana-3719 396 4 51 51 NUM cana-3719 396 5 , , PUNCT cana-3719 396 6 3460 3460 NUM cana-3719 396 7 - - SYM cana-3719 396 8 3474 3474 NUM cana-3719 396 9 . . PUNCT cana-3719 397 1 [ [ X cana-3719 397 2 44 44 NUM cana-3719 397 3 ] ] PUNCT cana-3719 397 4 . . PUNCT cana-3719 398 1 hedjazi hedjazi PROPN cana-3719 398 2 , , PUNCT cana-3719 398 3 m. m. NOUN cana-3719 398 4 a. a. PROPN cana-3719 398 5 , , PUNCT cana-3719 398 6 genc genc PROPN cana-3719 398 7 , , PUNCT cana-3719 398 8 y. y. NOUN cana-3719 398 9 ( ( PUNCT cana-3719 398 10 2021 2021 NUM cana-3719 398 11 ) ) PUNCT cana-3719 398 12 . . PUNCT cana-3719 399 1 efficient efficient ADJ cana-3719 399 2 texture texture ADJ cana-3719 399 3 - - PUNCT cana-3719 399 4 aware aware ADJ cana-3719 399 5 multi multi ADJ cana-3719 399 6 - - ADJ cana-3719 399 7 gan gan ADJ cana-3719 399 8 for for ADP cana-3719 399 9 image image NOUN cana-3719 399 10 inpainting inpainte VERB cana-3719 399 11 . . PUNCT cana-3719 400 1 knowledge knowledge NOUN cana-3719 400 2 - - PUNCT cana-3719 400 3 based base VERB cana-3719 400 4 systems system NOUN cana-3719 400 5 , , PUNCT cana-3719 400 6 217 217 NUM cana-3719 400 7 , , PUNCT cana-3719 400 8 106789 106789 NUM cana-3719 400 9 . . PUNCT cana-3719 401 1 [ [ X cana-3719 401 2 45 45 NUM cana-3719 401 3 ] ] PUNCT cana-3719 401 4 . . PUNCT cana-3719 402 1 liu liu PROPN cana-3719 402 2 , , PUNCT cana-3719 402 3 j. j. PROPN cana-3719 402 4 , , PUNCT cana-3719 402 5 gong gong PROPN cana-3719 402 6 , , PUNCT cana-3719 402 7 m. m. NOUN cana-3719 402 8 , , PUNCT cana-3719 402 9 tang tang PROPN cana-3719 402 10 , , PUNCT cana-3719 402 11 z. z. PROPN cana-3719 402 12 , , PUNCT cana-3719 402 13 qin qin PROPN cana-3719 402 14 , , PUNCT cana-3719 402 15 a. a. PROPN cana-3719 402 16 k. k. PROPN cana-3719 402 17 , , PUNCT cana-3719 402 18 li li PROPN cana-3719 402 19 , , PUNCT cana-3719 402 20 h. h. PROPN cana-3719 402 21 , , PUNCT cana-3719 402 22 jiang jiang PROPN cana-3719 402 23 , , PUNCT cana-3719 402 24 f. f. PROPN cana-3719 402 25 ( ( PUNCT cana-3719 402 26 2022 2022 NUM cana-3719 402 27 ) ) PUNCT cana-3719 402 28 . . PUNCT cana-3719 403 1 deep deep ADJ cana-3719 403 2 image image NOUN cana-3719 403 3 inpainting inpainte VERB cana-3719 403 4 with with ADP cana-3719 403 5 enhanced enhanced ADJ cana-3719 403 6 normalization normalization NOUN cana-3719 403 7 and and CCONJ cana-3719 403 8 contextual contextual ADJ cana-3719 403 9 attention attention NOUN cana-3719 403 10 . . PUNCT cana-3719 404 1 ieee ieee NOUN cana-3719 404 2 transactions transaction NOUN cana-3719 404 3 on on ADP cana-3719 404 4 circuits circuit NOUN cana-3719 404 5 and and CCONJ cana-3719 404 6 systems system NOUN cana-3719 404 7 for for ADP cana-3719 404 8 video video NOUN cana-3719 404 9 technology technology NOUN cana-3719 404 10 , , PUNCT cana-3719 404 11 32(10 32(10 NUM cana-3719 404 12 ) ) PUNCT cana-3719 404 13 , , PUNCT cana-3719 404 14 6599 6599 NUM cana-3719 404 15 - - SYM cana-3719 404 16 6614 6614 NUM cana-3719 404 17 . . PUNCT cana-3719 405 1 [ [ X cana-3719 405 2 46 46 NUM cana-3719 405 3 ] ] PUNCT cana-3719 405 4 . . PUNCT cana-3719 406 1 liu liu PROPN cana-3719 406 2 , , PUNCT cana-3719 406 3 h. h. PROPN cana-3719 406 4 , , PUNCT cana-3719 406 5 jiang jiang PROPN cana-3719 406 6 , , PUNCT cana-3719 406 7 b. b. PROPN cana-3719 406 8 , , PUNCT cana-3719 406 9 xiao xiao PROPN cana-3719 406 10 , , PUNCT cana-3719 406 11 y. y. PROPN cana-3719 406 12 , , PUNCT cana-3719 406 13 yang yang PROPN cana-3719 406 14 , , PUNCT cana-3719 406 15 c. c. PROPN cana-3719 406 16 ( ( PUNCT cana-3719 406 17 2019 2019 NUM cana-3719 406 18 ) ) PUNCT cana-3719 406 19 . . PUNCT cana-3719 407 1 coherent coherent ADJ cana-3719 407 2 semantic semantic ADJ cana-3719 407 3 attention attention NOUN cana-3719 407 4 for for ADP cana-3719 407 5 image image NOUN cana-3719 407 6 inpainting inpainting NOUN cana-3719 407 7 . . PUNCT cana-3719 408 1 in in ADP cana-3719 408 2 proceedings proceeding NOUN cana-3719 408 3 of of ADP cana-3719 408 4 the the DET cana-3719 408 5 ieee ieee NOUN cana-3719 408 6 / / SYM cana-3719 408 7 cvf cvf NOUN cana-3719 408 8 international international ADJ cana-3719 408 9 conference conference NOUN cana-3719 408 10 on on ADP cana-3719 408 11 computer computer NOUN cana-3719 408 12 vision vision NOUN cana-3719 408 13 ( ( PUNCT cana-3719 408 14 pp pp ADJ cana-3719 408 15 . . PUNCT cana-3719 408 16 4170 4170 NUM cana-3719 408 17 - - SYM cana-3719 408 18 4179 4179 NUM cana-3719 408 19 ) ) PUNCT cana-3719 408 20 . . PUNCT cana-3719 409 1 [ [ X cana-3719 409 2 47 47 NUM cana-3719 409 3 ] ] PUNCT cana-3719 409 4 . . PUNCT cana-3719 410 1 li li PROPN cana-3719 410 2 , , PUNCT cana-3719 410 3 c. c. PROPN cana-3719 410 4 , , PUNCT cana-3719 410 5 xu xu PROPN cana-3719 410 6 , , PUNCT cana-3719 410 7 d. d. PROPN cana-3719 410 8 , , PUNCT cana-3719 410 9 zhang zhang PROPN cana-3719 410 10 , , PUNCT cana-3719 410 11 h. h. PROPN cana-3719 410 12 ( ( PUNCT cana-3719 410 13 2024 2024 NUM cana-3719 410 14 ) ) PUNCT cana-3719 410 15 . . PUNCT cana-3719 411 1 multi‐stage multi‐stage NOUN cana-3719 411 2 image image NOUN cana-3719 411 3 inpainting inpainte VERB cana-3719 411 4 using use VERB cana-3719 411 5 improved improve VERB cana-3719 411 6 partial partial ADJ cana-3719 411 7 convolutions convolution NOUN cana-3719 411 8 . . PUNCT cana-3719 412 1 iet iet PROPN cana-3719 412 2 image image NOUN cana-3719 412 3 processing processing NOUN cana-3719 412 4 . . PUNCT cana-3719 413 1 [ [ X cana-3719 413 2 48 48 NUM cana-3719 413 3 ] ] PUNCT cana-3719 413 4 . . PUNCT cana-3719 414 1 chen chen PROPN cana-3719 414 2 , , PUNCT cana-3719 414 3 y. y. PROPN cana-3719 414 4 , , PUNCT cana-3719 414 5 xia xia PROPN cana-3719 414 6 , , PUNCT cana-3719 414 7 r. r. PROPN cana-3719 414 8 , , PUNCT cana-3719 414 9 yang yang PROPN cana-3719 414 10 , , PUNCT cana-3719 414 11 k. k. PROPN cana-3719 414 12 , , PUNCT cana-3719 414 13 zou zou PROPN cana-3719 414 14 , , PUNCT cana-3719 414 15 k. k. PROPN cana-3719 414 16 ( ( PUNCT cana-3719 414 17 2024 2024 NUM cana-3719 414 18 ) ) PUNCT cana-3719 414 19 . . PUNCT cana-3719 415 1 dnnam dnnam NOUN cana-3719 415 2 : : PUNCT cana-3719 415 3 image image NOUN cana-3719 415 4 inpainting inpainte VERB cana-3719 415 5 algorithm algorithm NOUN cana-3719 415 6 via via ADP cana-3719 415 7 deep deep ADJ cana-3719 415 8 neural neural ADJ cana-3719 415 9 networks network NOUN cana-3719 415 10 and and CCONJ cana-3719 415 11 attention attention NOUN cana-3719 415 12 mechanism mechanism NOUN cana-3719 415 13 . . PUNCT cana-3719 416 1 applied apply VERB cana-3719 416 2 soft soft ADJ cana-3719 416 3 computing computing NOUN cana-3719 416 4 , , PUNCT cana-3719 416 5 154 154 NUM cana-3719 416 6 , , PUNCT cana-3719 416 7 111392 111392 NUM cana-3719 416 8 . . PUNCT cana-3719 417 1 [ [ X cana-3719 417 2 49 49 NUM cana-3719 417 3 ] ] PUNCT cana-3719 417 4 . . PUNCT cana-3719 418 1 y. y. PROPN cana-3719 418 2 zeng zeng PROPN cana-3719 418 3 , , PUNCT cana-3719 418 4 j. j. PROPN cana-3719 418 5 fu fu PROPN cana-3719 418 6 , , PUNCT cana-3719 418 7 h. h. PROPN cana-3719 418 8 chao chao PROPN cana-3719 418 9 , , PUNCT cana-3719 418 10 and and CCONJ cana-3719 418 11 b. b. PROPN cana-3719 418 12 guo guo PROPN cana-3719 418 13 , , PUNCT cana-3719 418 14 β€˜ ' PUNCT cana-3719 418 15 β€˜ ' PUNCT cana-3719 418 16 learning learn VERB cana-3719 418 17 pyramid pyramid NOUN cana-3719 418 18 -context -context PROPN cana-3719 418 19 encoder encoder NOUN cana-3719 418 20 network network NOUN cana-3719 418 21 for for ADP cana-3719 418 22 high high ADJ cana-3719 418 23 - - PUNCT cana-3719 418 24 quality quality NOUN cana-3719 418 25 image image NOUN cana-3719 418 26 inpainting inpainte VERB cana-3719 418 27 , , PUNCT cana-3719 418 28 ’’ ’' PUNCT cana-3719 418 29 in in ADP cana-3719 418 30 proc proc NOUN cana-3719 418 31 . . PUNCT cana-3719 418 32 ieee ieee NOUN cana-3719 418 33 / / SYM cana-3719 418 34 cvf cvf NOUN cana-3719 418 35 conf conf NOUN cana-3719 418 36 . . PUNCT cana-3719 419 1 comput comput NOUN cana-3719 419 2 . . PUNCT cana-3719 420 1 vis vis X cana-3719 420 2 . . X cana-3719 420 3 pattern pattern NOUN cana-3719 420 4 recognit recognit VERB cana-3719 420 5 . . PUNCT cana-3719 421 1 ( ( PUNCT cana-3719 421 2 cvpr cvpr NOUN cana-3719 421 3 ) ) PUNCT cana-3719 421 4 , , PUNCT cana-3719 421 5 jun jun PROPN cana-3719 421 6 . . PROPN cana-3719 421 7 2019 2019 NUM cana-3719 421 8 , , PUNCT cana-3719 421 9 pp pp ADJ cana-3719 421 10 . . PUNCT cana-3719 422 1 1486 1486 NUM cana-3719 422 2 – – PUNCT cana-3719 422 3 1494 1494 NUM cana-3719 423 1 [ [ NOUN cana-3719 423 2 50 50 NUM cana-3719 423 3 ] ] PUNCT cana-3719 423 4 . . PUNCT cana-3719 424 1 g. g. PROPN cana-3719 424 2 liu liu PROPN cana-3719 424 3 , , PUNCT cana-3719 424 4 f. f. PROPN cana-3719 424 5 a. a. PROPN cana-3719 424 6 reda reda PROPN cana-3719 424 7 , , PUNCT cana-3719 424 8 k. k. PROPN cana-3719 424 9 j. j. PROPN cana-3719 424 10 shih shih PROPN cana-3719 424 11 , , PUNCT cana-3719 424 12 t.-c t.-c PROPN cana-3719 424 13 . . PUNCT cana-3719 425 1 wang wang PROPN cana-3719 425 2 , , PUNCT cana-3719 425 3 a. a. NOUN cana-3719 425 4 tao tao PROPN cana-3719 425 5 , , PUNCT cana-3719 425 6 and and CCONJ cana-3719 425 7 b. b. PROPN cana-3719 425 8 catanzaro catanzaro PROPN cana-3719 425 9 , , PUNCT cana-3719 425 10 β€˜ ' PUNCT cana-3719 425 11 β€˜ ' PUNCT cana-3719 425 12 image image NOUN cana-3719 425 13 inpainting inpainte VERB cana-3719 425 14 for for ADP cana-3719 425 15 irregular irregular ADJ cana-3719 425 16 holes hole NOUN cana-3719 425 17 using use VERB cana-3719 425 18 partial partial ADJ cana-3719 425 19 convolutions convolution NOUN cana-3719 425 20 , , PUNCT cana-3719 425 21 ’’ ’' PUNCT cana-3719 425 22 in in ADP cana-3719 425 23 proc proc NOUN cana-3719 425 24 . . PUNCT cana-3719 426 1 eur eur PROPN cana-3719 426 2 . . PUNCT cana-3719 426 3 conf conf PROPN cana-3719 426 4 . . PUNCT cana-3719 427 1 comput comput NOUN cana-3719 427 2 . . PUNCT cana-3719 428 1 vis vis X cana-3719 428 2 . . PUNCT cana-3719 428 3 ( ( PUNCT cana-3719 428 4 eccv eccv ADV cana-3719 428 5 ) ) PUNCT cana-3719 428 6 , , PUNCT cana-3719 428 7 2018 2018 NUM cana-3719 428 8 , , PUNCT cana-3719 428 9 pp pp ADJ cana-3719 428 10 . . PUNCT cana-3719 429 1 85–100 85–100 NOUN cana-3719 429 2 . . PUNCT cana-3719 430 1 [ [ X cana-3719 430 2 51 51 NUM cana-3719 430 3 ] ] PUNCT cana-3719 430 4 . . PUNCT cana-3719 431 1 x. x. PROPN cana-3719 431 2 hong hong PROPN cana-3719 431 3 , , PUNCT cana-3719 431 4 p. p. PROPN cana-3719 431 5 xiong xiong PROPN cana-3719 431 6 , , PUNCT cana-3719 431 7 r. r. PROPN cana-3719 431 8 ji ji PROPN cana-3719 431 9 , , PUNCT cana-3719 431 10 and and CCONJ cana-3719 431 11 h. h. PROPN cana-3719 431 12 fan fan PROPN cana-3719 431 13 , , PUNCT cana-3719 431 14 β€˜ ' PUNCT cana-3719 431 15 β€˜ ' PUNCT cana-3719 431 16 deep deep ADJ cana-3719 431 17 fusion fusion NOUN cana-3719 431 18 network network NOUN cana-3719 431 19 for for ADP cana-3719 431 20 image image NOUN cana-3719 431 21 completion completion NOUN cana-3719 431 22 , , PUNCT cana-3719 431 23 ’’ ’' PUNCT cana-3719 431 24 in in ADP cana-3719 431 25 proc proc NOUN cana-3719 431 26 . . PUNCT cana-3719 432 1 27th 27th PROPN cana-3719 432 2 acm acm PROPN cana-3719 432 3 int int PROPN cana-3719 432 4 . . PUNCT cana-3719 432 5 conf conf PROPN cana-3719 432 6 . . PUNCT cana-3719 433 1 multimedia multimedia PROPN cana-3719 433 2 , , PUNCT cana-3719 433 3 oct oct PROPN cana-3719 433 4 . . PROPN cana-3719 433 5 2019 2019 NUM cana-3719 433 6 , , PUNCT cana-3719 433 7 pp pp ADJ cana-3719 433 8 . . PUNCT cana-3719 434 1 2033–2042 2033–2042 NUM cana-3719 434 2 . . PUNCT cana-3719 435 1 [ [ X cana-3719 435 2 52 52 NUM cana-3719 435 3 ] ] PUNCT cana-3719 435 4 . . PUNCT cana-3719 436 1 z. z. PROPN cana-3719 436 2 yan yan PROPN cana-3719 436 3 , , PUNCT cana-3719 436 4 x. x. PROPN cana-3719 436 5 li li PROPN cana-3719 436 6 , , PUNCT cana-3719 436 7 m. m. PROPN cana-3719 436 8 li li PROPN cana-3719 436 9 , , PUNCT cana-3719 436 10 w. w. PROPN cana-3719 436 11 zuo zuo PROPN cana-3719 436 12 , , PUNCT cana-3719 436 13 and and CCONJ cana-3719 436 14 s. s. PROPN cana-3719 436 15 shan shan PROPN cana-3719 436 16 , , PUNCT cana-3719 436 17 β€˜ ' PUNCT cana-3719 436 18 β€˜ ' PUNCT cana-3719 436 19 shift shift NOUN cana-3719 436 20 - - PUNCT cana-3719 436 21 net net NOUN cana-3719 436 22 : : PUNCT cana-3719 436 23 image image NOUN cana-3719 436 24 inpainting inpainte VERB cana-3719 436 25 via via ADP cana-3719 436 26 deep deep ADJ cana-3719 436 27 feature feature NOUN cana-3719 436 28 rearrangement rearrangement NOUN cana-3719 436 29 , , PUNCT cana-3719 436 30 ’’ ’' PUNCT cana-3719 436 31 in in ADP cana-3719 436 32 proc proc NOUN cana-3719 436 33 . . PUNCT cana-3719 437 1 eur eur PROPN cana-3719 437 2 . . PUNCT cana-3719 437 3 conf conf PROPN cana-3719 437 4 . . PUNCT cana-3719 438 1 comput comput NOUN cana-3719 438 2 . . PUNCT cana-3719 439 1 vis vis X cana-3719 439 2 . . PUNCT cana-3719 439 3 ( ( PUNCT cana-3719 439 4 eccv eccv ADV cana-3719 439 5 ) ) PUNCT cana-3719 439 6 , , PUNCT cana-3719 439 7 2018 2018 NUM cana-3719 439 8 , , PUNCT cana-3719 439 9 pp pp ADJ cana-3719 439 10 . . PUNCT cana-3719 440 1 1–17 1–17 NOUN cana-3719 440 2 . . PUNCT cana-3719 441 1 [ [ X cana-3719 441 2 53 53 NUM cana-3719 441 3 ] ] PUNCT cana-3719 441 4 . . PUNCT cana-3719 442 1 liu liu PROPN cana-3719 442 2 , , PUNCT cana-3719 442 3 l. l. PROPN cana-3719 442 4 , , PUNCT cana-3719 442 5 liu liu PROPN cana-3719 442 6 , , PUNCT cana-3719 442 7 y. y. PROPN cana-3719 442 8 ( ( PUNCT cana-3719 442 9 2022 2022 NUM cana-3719 442 10 ) ) PUNCT cana-3719 442 11 . . PUNCT cana-3719 443 1 load load NOUN cana-3719 443 2 image image NOUN cana-3719 443 3 inpainting inpainte VERB cana-3719 443 4 : : PUNCT cana-3719 443 5 an an DET cana-3719 443 6 improved improved ADJ cana-3719 443 7 u u NOUN cana-3719 443 8 - - ADJ cana-3719 443 9 net net ADJ cana-3719 443 10 based base VERB cana-3719 443 11 load load NOUN cana-3719 443 12 missing miss VERB cana-3719 443 13 data datum NOUN cana-3719 443 14 recovery recovery NOUN cana-3719 443 15 method method NOUN cana-3719 443 16 . . PUNCT cana-3719 444 1 applied apply VERB cana-3719 444 2 energy energy NOUN cana-3719 444 3 , , PUNCT cana-3719 444 4 327 327 NUM cana-3719 444 5 , , PUNCT cana-3719 444 6 119988 119988 NUM cana-3719 444 7 . . PUNCT cana-3719 445 1 [ [ X cana-3719 445 2 54 54 NUM cana-3719 445 3 ] ] PUNCT cana-3719 445 4 . . PUNCT cana-3719 446 1 hou hou PROPN cana-3719 446 2 , , PUNCT cana-3719 446 3 y. y. PROPN cana-3719 446 4 , , PUNCT cana-3719 446 5 ma ma PROPN cana-3719 446 6 , , PUNCT cana-3719 446 7 x. x. PROPN cana-3719 446 8 , , PUNCT cana-3719 446 9 zhang zhang PROPN cana-3719 446 10 , , PUNCT cana-3719 446 11 j. j. PROPN cana-3719 446 12 , , PUNCT cana-3719 446 13 guo guo PROPN cana-3719 446 14 , , PUNCT cana-3719 446 15 c. c. PROPN cana-3719 446 16 ( ( PUNCT cana-3719 446 17 2024 2024 NUM cana-3719 446 18 ) ) PUNCT cana-3719 446 19 . . PUNCT cana-3719 447 1 symmetric symmetric ADJ cana-3719 447 2 connected connect VERB cana-3719 447 3 u u NOUN cana-3719 447 4 - - NOUN cana-3719 447 5 net net ADJ cana-3719 447 6 with with ADP cana-3719 447 7 multi multi ADJ cana-3719 447 8 - - ADJ cana-3719 447 9 head head ADJ cana-3719 447 10 self self NOUN cana-3719 447 11 attention attention NOUN cana-3719 447 12 ( ( PUNCT cana-3719 447 13 mhsa mhsa NOUN cana-3719 447 14 ) ) PUNCT cana-3719 447 15 and and CCONJ cana-3719 447 16 wgan wgan VERB cana-3719 447 17 for for ADP cana-3719 447 18 image image NOUN cana-3719 447 19 inpainting inpainte VERB cana-3719 447 20 . . PUNCT cana-3719 447 21 symmetry symmetry NOUN cana-3719 447 22 , , PUNCT cana-3719 447 23 16(11 16(11 NUM cana-3719 447 24 ) ) PUNCT cana-3719 447 25 , , PUNCT cana-3719 447 26 1423 1423 NUM cana-3719 447 27 . . PUNCT cana-3719 448 1 [ [ X cana-3719 448 2 55 55 NUM cana-3719 448 3 ] ] PUNCT cana-3719 448 4 . . PUNCT cana-3719 449 1 quan quan PROPN cana-3719 449 2 , , PUNCT cana-3719 449 3 w. w. PROPN cana-3719 449 4 , , PUNCT cana-3719 449 5 zhang zhang PROPN cana-3719 449 6 , , PUNCT cana-3719 449 7 r. r. PROPN cana-3719 449 8 , , PUNCT cana-3719 449 9 zhang zhang PROPN cana-3719 449 10 , , PUNCT cana-3719 449 11 y. y. PROPN cana-3719 449 12 , , PUNCT cana-3719 449 13 li li PROPN cana-3719 449 14 , , PUNCT cana-3719 449 15 z. z. PROPN cana-3719 449 16 , , PUNCT cana-3719 449 17 wang wang PROPN cana-3719 449 18 , , PUNCT cana-3719 449 19 j. j. PROPN cana-3719 449 20 , , PUNCT cana-3719 449 21 yan yan PROPN cana-3719 449 22 , , PUNCT cana-3719 449 23 d. d. PROPN cana-3719 449 24 m. m. PROPN cana-3719 449 25 ( ( PUNCT cana-3719 449 26 2022 2022 NUM cana-3719 449 27 ) ) PUNCT cana-3719 449 28 . . PUNCT cana-3719 450 1 image image NOUN cana-3719 450 2 inpainting inpainte VERB cana-3719 450 3 with with ADP cana-3719 450 4 local local ADJ cana-3719 450 5 and and CCONJ cana-3719 450 6 global global ADJ cana-3719 450 7 refinement refinement NOUN cana-3719 450 8 . . PUNCT cana-3719 451 1 ieee ieee NOUN cana-3719 451 2 transactions transaction NOUN cana-3719 451 3 on on ADP cana-3719 451 4 image image NOUN cana-3719 451 5 processing processing NOUN cana-3719 451 6 , , PUNCT cana-3719 451 7 31 31 NUM cana-3719 451 8 , , PUNCT cana-3719 451 9 2405 2405 NUM cana-3719 451 10 - - SYM cana-3719 451 11 2420 2420 NUM cana-3719 451 12 . . PUNCT cana-3719 452 1 [ [ X cana-3719 452 2 56 56 NUM cana-3719 452 3 ] ] PUNCT cana-3719 452 4 . . PUNCT cana-3719 453 1 liu liu PROPN cana-3719 453 2 , , PUNCT cana-3719 453 3 j. j. PROPN cana-3719 453 4 , , PUNCT cana-3719 453 5 gong gong PROPN cana-3719 453 6 , , PUNCT cana-3719 453 7 m. m. NOUN cana-3719 453 8 , , PUNCT cana-3719 453 9 gao gao PROPN cana-3719 453 10 , , PUNCT cana-3719 453 11 y. y. PROPN cana-3719 453 12 , , PUNCT cana-3719 453 13 lu lu PROPN cana-3719 453 14 , , PUNCT cana-3719 453 15 y. y. PROPN cana-3719 453 16 , , PUNCT cana-3719 453 17 li li PROPN cana-3719 453 18 , , PUNCT cana-3719 453 19 h. h. PROPN cana-3719 453 20 ( ( PUNCT cana-3719 453 21 2024 2024 NUM cana-3719 453 22 ) ) PUNCT cana-3719 453 23 . . PUNCT cana-3719 454 1 bidirectional bidirectional ADJ cana-3719 454 2 interaction interaction NOUN cana-3719 454 3 of of ADP cana-3719 454 4 cnn cnn PROPN cana-3719 454 5 and and CCONJ cana-3719 454 6 transformer transformer NOUN cana-3719 454 7 for for ADP cana-3719 454 8 image image NOUN cana-3719 454 9 inpainting inpainte VERB cana-3719 454 10 . . PUNCT cana-3719 455 1 knowledge knowledge NOUN cana-3719 455 2 - - PUNCT cana-3719 455 3 based base VERB cana-3719 455 4 systems system NOUN cana-3719 455 5 , , PUNCT cana-3719 455 6 112046 112046 NUM cana-3719 455 7 . . PUNCT cana-3719 456 1 [ [ X cana-3719 456 2 57 57 NUM cana-3719 456 3 ] ] PUNCT cana-3719 456 4 . . PUNCT cana-3719 457 1 chen chen PROPN cana-3719 457 2 , , PUNCT cana-3719 457 3 m. m. NOUN cana-3719 457 4 , , PUNCT cana-3719 457 5 zang zang PROPN cana-3719 457 6 , , PUNCT cana-3719 457 7 s. s. PROPN cana-3719 457 8 , , PUNCT cana-3719 457 9 ai ai VERB cana-3719 457 10 , , PUNCT cana-3719 457 11 z. z. PROPN cana-3719 457 12 , , PUNCT cana-3719 457 13 chi chi PROPN cana-3719 457 14 , , PUNCT cana-3719 457 15 j. j. PROPN cana-3719 457 16 , , PUNCT cana-3719 457 17 yang yang PROPN cana-3719 457 18 , , PUNCT cana-3719 457 19 g. g. PROPN cana-3719 457 20 , , PUNCT cana-3719 457 21 chen chen PROPN cana-3719 457 22 , , PUNCT cana-3719 457 23 c. c. PROPN cana-3719 457 24 , , PUNCT cana-3719 457 25 yu yu PROPN cana-3719 457 26 , , PUNCT cana-3719 457 27 t. t. PROPN cana-3719 457 28 ( ( PUNCT cana-3719 457 29 2023 2023 NUM cana-3719 457 30 ) ) PUNCT cana-3719 457 31 . . PUNCT cana-3719 458 1 rfa rfa NOUN cana-3719 458 2 - - PUNCT cana-3719 458 3 net net NOUN cana-3719 458 4 : : PUNCT cana-3719 458 5 residual residual ADJ cana-3719 458 6 feature feature NOUN cana-3719 458 7 attention attention NOUN cana-3719 458 8 network network NOUN cana-3719 458 9 for for ADP cana-3719 458 10 fine fine ADV cana-3719 458 11 - - PUNCT cana-3719 458 12 grained grain VERB cana-3719 458 13 image image NOUN cana-3719 458 14 inpainting inpainte VERB cana-3719 458 15 . . PUNCT cana-3719 459 1 engineering engineering NOUN cana-3719 459 2 applications application NOUN cana-3719 459 3 of of ADP cana-3719 459 4 artificial artificial ADJ cana-3719 459 5 intelligence intelligence NOUN cana-3719 459 6 , , PUNCT cana-3719 459 7 119 119 NUM cana-3719 459 8 , , PUNCT cana-3719 459 9 105814 105814 NUM cana-3719 459 10 . . PUNCT cana-3719 460 1 [ [ X cana-3719 460 2 58 58 NUM cana-3719 460 3 ] ] PUNCT cana-3719 460 4 . . PUNCT cana-3719 461 1 meng meng PROPN cana-3719 461 2 , , PUNCT cana-3719 461 3 j. j. PROPN cana-3719 461 4 , , PUNCT cana-3719 461 5 liu liu PROPN cana-3719 461 6 , , PUNCT cana-3719 461 7 w. w. PROPN cana-3719 461 8 , , PUNCT cana-3719 461 9 shi shi PROPN cana-3719 461 10 , , PUNCT cana-3719 461 11 c. c. PROPN cana-3719 461 12 , , PUNCT cana-3719 461 13 li li PROPN cana-3719 461 14 , , PUNCT cana-3719 461 15 z. z. PROPN cana-3719 461 16 , , PUNCT cana-3719 461 17 liu liu PROPN cana-3719 461 18 , , PUNCT cana-3719 461 19 c. c. PROPN cana-3719 461 20 ( ( PUNCT cana-3719 461 21 2024 2024 NUM cana-3719 461 22 ) ) PUNCT cana-3719 461 23 . . PUNCT cana-3719 462 1 degression degression NOUN cana-3719 462 2 receptive receptive ADJ cana-3719 462 3 field field NOUN cana-3719 462 4 network network NOUN cana-3719 462 5 for for ADP cana-3719 462 6 image image NOUN cana-3719 462 7 inpainting inpainte VERB cana-3719 462 8 . . PUNCT cana-3719 463 1 engineering engineering NOUN cana-3719 463 2 applications application NOUN cana-3719 463 3 of of ADP cana-3719 463 4 artificial artificial ADJ cana-3719 463 5 intelligence intelligence NOUN cana-3719 463 6 , , PUNCT cana-3719 463 7 138 138 NUM cana-3719 463 8 , , PUNCT cana-3719 463 9 109397 109397 NUM cana-3719 463 10 . . PUNCT cana-3719 464 1 [ [ X cana-3719 464 2 59 59 NUM cana-3719 464 3 ] ] PUNCT cana-3719 464 4 . . PUNCT cana-3719 465 1 ma ma PROPN cana-3719 465 2 , , PUNCT cana-3719 465 3 x. x. PROPN cana-3719 465 4 , , PUNCT cana-3719 465 5 zhou zhou PROPN cana-3719 465 6 , , PUNCT cana-3719 465 7 x. x. PROPN cana-3719 465 8 , , PUNCT cana-3719 465 9 huang huang PROPN cana-3719 465 10 , , PUNCT cana-3719 465 11 h. h. PROPN cana-3719 465 12 , , PUNCT cana-3719 465 13 jia jia PROPN cana-3719 465 14 , , PUNCT cana-3719 465 15 g. g. PROPN cana-3719 465 16 , , PUNCT cana-3719 465 17 wang wang PROPN cana-3719 465 18 , , PUNCT cana-3719 465 19 y. y. PROPN cana-3719 465 20 , , PUNCT cana-3719 465 21 chen chen PROPN cana-3719 465 22 , , PUNCT cana-3719 465 23 x. x. PROPN cana-3719 465 24 , , PUNCT cana-3719 465 25 chen chen PROPN cana-3719 465 26 , , PUNCT cana-3719 465 27 c. c. PROPN cana-3719 465 28 ( ( PUNCT cana-3719 465 29 2024 2024 NUM cana-3719 465 30 ) ) PUNCT cana-3719 465 31 . . PUNCT cana-3719 466 1 uncertainty uncertainty NOUN cana-3719 466 2 -aware -aware NOUN cana-3719 466 3 image image NOUN cana-3719 466 4 inpainting inpainte VERB cana-3719 466 5 with with ADP cana-3719 466 6 adaptive adaptive ADJ cana-3719 466 7 feedback feedback NOUN cana-3719 466 8 network network NOUN cana-3719 466 9 . . PUNCT cana-3719 467 1 expert expert NOUN cana-3719 467 2 systems system NOUN cana-3719 467 3 with with ADP cana-3719 467 4 applications application NOUN cana-3719 467 5 , , PUNCT cana-3719 467 6 235 235 NUM cana-3719 467 7 , , PUNCT cana-3719 467 8 121148 121148 NUM cana-3719 467 9 . . PUNCT