id sid tid token lemma pos bracis-19071 1 1 iris iris PROPN bracis-19071 1 2 - - PUNCT bracis-19071 1 3 cv cv PROPN bracis-19071 1 4 : : PUNCT bracis-19071 1 5 classifying classify VERB bracis-19071 1 6 iris iris NOUN bracis-19071 1 7 flowers flower NOUN bracis-19071 1 8 is be AUX bracis-19071 1 9 not not PART bracis-19071 1 10 as as ADV bracis-19071 1 11 easy easy ADJ bracis-19071 1 12 as as SCONJ bracis-19071 1 13 you you PRON bracis-19071 1 14 thought think VERB bracis-19071 1 15 | | ADV bracis-19071 1 16 springer springer NOUN bracis-19071 1 17 nature nature PROPN bracis-19071 1 18 link link PROPN bracis-19071 1 19 ( ( PUNCT bracis-19071 1 20 formerly formerly ADV bracis-19071 1 21 springerlink springerlink NOUN bracis-19071 1 22 ) ) PUNCT bracis-19071 1 23 skip skip VERB bracis-19071 1 24 to to ADP bracis-19071 1 25 main main ADJ bracis-19071 1 26 content content NOUN bracis-19071 1 27 advertisement advertisement NOUN bracis-19071 1 28 log log NOUN bracis-19071 1 29 in in ADP bracis-19071 1 30 menu menu NOUN bracis-19071 1 31 find find VERB bracis-19071 1 32 a a DET bracis-19071 1 33 journal journal NOUN bracis-19071 1 34 publish publish VERB bracis-19071 1 35 with with ADP bracis-19071 1 36 us we PRON bracis-19071 1 37 track track VERB bracis-19071 1 38 your your PRON bracis-19071 1 39 research research NOUN bracis-19071 1 40 search search NOUN bracis-19071 1 41 cart cart NOUN bracis-19071 1 42 home home NOUN bracis-19071 1 43 intelligent intelligent ADJ bracis-19071 1 44 systems system NOUN bracis-19071 1 45 conference conference NOUN bracis-19071 1 46 paper paper NOUN bracis-19071 1 47 iris iris PROPN bracis-19071 1 48 - - PUNCT bracis-19071 1 49 cv cv PROPN bracis-19071 1 50 : : PUNCT bracis-19071 1 51 classifying classify VERB bracis-19071 1 52 iris iris NOUN bracis-19071 1 53 flowers flower NOUN bracis-19071 1 54 is be AUX bracis-19071 1 55 not not PART bracis-19071 1 56 as as ADV bracis-19071 1 57 easy easy ADJ bracis-19071 1 58 as as SCONJ bracis-19071 1 59 you you PRON bracis-19071 1 60 thought think VERB bracis-19071 1 61 conference conference NOUN bracis-19071 1 62 paper paper NOUN bracis-19071 1 63 first first ADV bracis-19071 1 64 online online ADV bracis-19071 1 65 : : PUNCT bracis-19071 1 66 28 28 NUM bracis-19071 1 67 november november NOUN bracis-19071 1 68 2021 2021 NUM bracis-19071 1 69 pp pp ADP bracis-19071 1 70 254–264 254–264 NUM bracis-19071 1 71 cite cite VERB bracis-19071 1 72 this this DET bracis-19071 1 73 conference conference NOUN bracis-19071 1 74 paper paper NOUN bracis-19071 1 75 access access NOUN bracis-19071 1 76 provided provide VERB bracis-19071 1 77 by by ADP bracis-19071 1 78 university university PROPN bracis-19071 1 79 of of ADP bracis-19071 1 80 notre notre PROPN bracis-19071 1 81 dame dame PROPN bracis-19071 1 82 hesburgh hesburgh PROPN bracis-19071 1 83 library library PROPN bracis-19071 1 84 download download PROPN bracis-19071 1 85 book book NOUN bracis-19071 1 86 pdf pdf PROPN bracis-19071 1 87 download download NOUN bracis-19071 1 88 book book NOUN bracis-19071 1 89 epub epub PROPN bracis-19071 1 90 intelligent intelligent ADJ bracis-19071 1 91 systems system NOUN bracis-19071 1 92 ( ( PUNCT bracis-19071 1 93 bracis bracis NOUN bracis-19071 1 94 2021 2021 NUM bracis-19071 1 95 ) ) PUNCT bracis-19071 1 96 iris iris NOUN bracis-19071 1 97 - - PUNCT bracis-19071 1 98 cv cv PROPN bracis-19071 1 99 : : PUNCT bracis-19071 1 100 classifying classify VERB bracis-19071 1 101 iris iris NOUN bracis-19071 1 102 flowers flower NOUN bracis-19071 1 103 is be AUX bracis-19071 1 104 not not PART bracis-19071 1 105 as as ADV bracis-19071 1 106 easy easy ADJ bracis-19071 1 107 as as SCONJ bracis-19071 1 108 you you PRON bracis-19071 1 109 thought think VERB bracis-19071 1 110 download download NOUN bracis-19071 1 111 book book NOUN bracis-19071 1 112 pdf pdf PROPN bracis-19071 1 113 download download NOUN bracis-19071 1 114 book book PROPN bracis-19071 1 115 epub epub PROPN bracis-19071 1 116 itamar itamar PROPN bracis-19071 1 117 de de PROPN bracis-19071 1 118 paiva paiva PROPN bracis-19071 1 119 rocha rocha PROPN bracis-19071 1 120 filho10 filho10 PROPN bracis-19071 1 121 , , PUNCT bracis-19071 1 122 joão joão PROPN bracis-19071 1 123 pedro pedro PROPN bracis-19071 1 124 vasconcelos vasconcelos PROPN bracis-19071 1 125 teixeira10 teixeira10 PROPN bracis-19071 1 126 , , PUNCT bracis-19071 1 127 joão joão PROPN bracis-19071 1 128 wallace wallace PROPN bracis-19071 1 129 lucena lucena PROPN bracis-19071 1 130 lins10 lins10 PROPN bracis-19071 1 131 , , PUNCT bracis-19071 1 132 felipe felipe PROPN bracis-19071 1 133 honorato honorato PROPN bracis-19071 1 134 de de X bracis-19071 1 135 sousa10 sousa10 NOUN bracis-19071 1 136 , , PUNCT bracis-19071 1 137 ana ana PROPN bracis-19071 1 138 clara clara PROPN bracis-19071 1 139 chaves chaves PROPN bracis-19071 1 140 sousa10 sousa10 PROPN bracis-19071 1 141 , , PUNCT bracis-19071 1 142 manuel manuel PROPN bracis-19071 1 143 ferreira ferreira PROPN bracis-19071 1 144 junior10 junior10 PROPN bracis-19071 1 145 , , PUNCT bracis-19071 1 146 thaís thaís NOUN bracis-19071 1 147 ramos11 ramos11 PROPN bracis-19071 1 148 , , PUNCT bracis-19071 1 149 cecília cecília PROPN bracis-19071 1 150 silva12 silva12 ADJ bracis-19071 1 151 , , PUNCT bracis-19071 1 152 thaís thaís NOUN bracis-19071 1 153 gaudencio gaudencio NOUN bracis-19071 1 154 do do AUX bracis-19071 1 155 rêgo10 rêgo10 VERB bracis-19071 1 156 , , PUNCT bracis-19071 1 157 yuri yuri PROPN bracis-19071 1 158 de de PROPN bracis-19071 1 159 almeida almeida PROPN bracis-19071 1 160 malheiros10 malheiros10 PROPN bracis-19071 1 161 & & CCONJ bracis-19071 1 162 … … PUNCT bracis-19071 1 163 telmo telmo PROPN bracis-19071 1 164 silva silva PROPN bracis-19071 1 165 filho10 filho10 PROPN bracis-19071 1 166     SPACE bracis-19071 1 167 show show NOUN bracis-19071 1 168 authors author NOUN bracis-19071 1 169 part part NOUN bracis-19071 1 170 of of ADP bracis-19071 1 171 the the DET bracis-19071 1 172 book book NOUN bracis-19071 1 173 series series NOUN bracis-19071 1 174 : : PUNCT bracis-19071 1 175 lecture lecture NOUN bracis-19071 1 176 notes note NOUN bracis-19071 1 177 in in ADP bracis-19071 1 178 computer computer NOUN bracis-19071 1 179 science science NOUN bracis-19071 1 180 ( ( PUNCT bracis-19071 1 181 ( ( PUNCT bracis-19071 1 182 lnai lnai ADJ bracis-19071 1 183 , , PUNCT bracis-19071 1 184 volume volume NOUN bracis-19071 1 185 13074 13074 NUM bracis-19071 1 186 ) ) PUNCT bracis-19071 1 187 ) ) PUNCT bracis-19071 1 188 included include VERB bracis-19071 1 189 in in ADP bracis-19071 1 190 the the DET bracis-19071 1 191 following follow VERB bracis-19071 1 192 conference conference NOUN bracis-19071 1 193 series series NOUN bracis-19071 1 194 : : PUNCT bracis-19071 1 195 brazilian brazilian ADJ bracis-19071 1 196 conference conference NOUN bracis-19071 1 197 on on ADP bracis-19071 1 198 intelligent intelligent ADJ bracis-19071 1 199 systems system NOUN bracis-19071 1 200 1275 1275 NUM bracis-19071 1 201 accesses access NOUN bracis-19071 1 202 abstract abstract ADV bracis-19071 1 203 the the DET bracis-19071 1 204 iris iris NOUN bracis-19071 1 205 flower flower NOUN bracis-19071 1 206 dataset dataset NOUN bracis-19071 1 207 is be AUX bracis-19071 1 208 a a DET bracis-19071 1 209 ubiquitous ubiquitous ADJ bracis-19071 1 210 benchmark benchmark NOUN bracis-19071 1 211 task task NOUN bracis-19071 1 212 in in ADP bracis-19071 1 213 machine machine NOUN bracis-19071 1 214 learning learn VERB bracis-19071 1 215 literature literature NOUN bracis-19071 1 216 . . PUNCT bracis-19071 2 1 with with ADP bracis-19071 2 2 its its PRON bracis-19071 2 3 150 150 NUM bracis-19071 2 4 instances instance NOUN bracis-19071 2 5 , , PUNCT bracis-19071 2 6 four four NUM bracis-19071 2 7 continuous continuous ADJ bracis-19071 2 8 features feature NOUN bracis-19071 2 9 , , PUNCT bracis-19071 2 10 and and CCONJ bracis-19071 2 11 three three NUM bracis-19071 2 12 balanced balanced ADJ bracis-19071 2 13 classes class NOUN bracis-19071 2 14 , , PUNCT bracis-19071 2 15 of of ADP bracis-19071 2 16 which which PRON bracis-19071 2 17 one one NOUN bracis-19071 2 18 is be AUX bracis-19071 2 19 linearly linearly ADV bracis-19071 2 20 separable separable ADJ bracis-19071 2 21 from from ADP bracis-19071 2 22 the the DET bracis-19071 2 23 others other NOUN bracis-19071 2 24 , , PUNCT bracis-19071 2 25 iris iris NOUN bracis-19071 2 26 is be AUX bracis-19071 2 27 generally generally ADV bracis-19071 2 28 considered consider VERB bracis-19071 2 29 an an DET bracis-19071 2 30 easy easy ADJ bracis-19071 2 31 problem problem NOUN bracis-19071 2 32 . . PUNCT bracis-19071 3 1 hence hence ADV bracis-19071 3 2 researchers researcher NOUN bracis-19071 3 3 usually usually ADV bracis-19071 3 4 rely rely VERB bracis-19071 3 5 on on ADP bracis-19071 3 6 other other ADJ bracis-19071 3 7 datasets dataset NOUN bracis-19071 3 8 when when SCONJ bracis-19071 3 9 they they PRON bracis-19071 3 10 need need VERB bracis-19071 3 11 more more ADV bracis-19071 3 12 challenging challenging ADJ bracis-19071 3 13 benchmarks benchmark NOUN bracis-19071 3 14 . . PUNCT bracis-19071 4 1 a a DET bracis-19071 4 2 similar similar ADJ bracis-19071 4 3 situation situation NOUN bracis-19071 4 4 happens happen VERB bracis-19071 4 5 with with ADP bracis-19071 4 6 computer computer NOUN bracis-19071 4 7 vision vision NOUN bracis-19071 4 8 datasets dataset NOUN bracis-19071 4 9 such such ADJ bracis-19071 4 10 as as ADP bracis-19071 4 11 mnist mnist NOUN bracis-19071 4 12 and and CCONJ bracis-19071 4 13 imagenet imagenet NOUN bracis-19071 4 14 , , PUNCT bracis-19071 4 15 which which PRON bracis-19071 4 16 have have AUX bracis-19071 4 17 been be AUX bracis-19071 4 18 widely widely ADV bracis-19071 4 19 explored explore VERB bracis-19071 4 20 . . PUNCT bracis-19071 5 1 the the DET bracis-19071 5 2 state state NOUN bracis-19071 5 3 of of ADP bracis-19071 5 4 the the DET bracis-19071 5 5 art art NOUN bracis-19071 5 6 models model NOUN bracis-19071 5 7 essentially essentially ADV bracis-19071 5 8 solves solve VERB bracis-19071 5 9 these these DET bracis-19071 5 10 problems problem NOUN bracis-19071 5 11 , , PUNCT bracis-19071 5 12 motivating motivate VERB bracis-19071 5 13 the the DET bracis-19071 5 14 search search NOUN bracis-19071 5 15 for for ADP bracis-19071 5 16 more more ADV bracis-19071 5 17 challenging challenging ADJ bracis-19071 5 18 tasks task NOUN bracis-19071 5 19 . . PUNCT bracis-19071 6 1 therefore therefore ADV bracis-19071 6 2 , , PUNCT bracis-19071 6 3 this this DET bracis-19071 6 4 paper paper NOUN bracis-19071 6 5 introduces introduce VERB bracis-19071 6 6 a a DET bracis-19071 6 7 new new ADJ bracis-19071 6 8 computer computer NOUN bracis-19071 6 9 vision vision NOUN bracis-19071 6 10 toy toy NOUN bracis-19071 6 11 dataset dataset NOUN bracis-19071 6 12 featuring feature VERB bracis-19071 6 13 iris iris NOUN bracis-19071 6 14 flowers flower NOUN bracis-19071 6 15 . . PUNCT bracis-19071 7 1 users user NOUN bracis-19071 7 2 of of ADP bracis-19071 7 3 a a DET bracis-19071 7 4 nature nature NOUN bracis-19071 7 5 photography photography NOUN bracis-19071 7 6 application application NOUN bracis-19071 7 7 took take VERB bracis-19071 7 8 the the DET bracis-19071 7 9 pictures picture NOUN bracis-19071 7 10 , , PUNCT bracis-19071 7 11 thus thus ADV bracis-19071 7 12 they they PRON bracis-19071 7 13 include include VERB bracis-19071 7 14 noisy noisy ADJ bracis-19071 7 15 background background NOUN bracis-19071 7 16 information information NOUN bracis-19071 7 17 . . PUNCT bracis-19071 8 1 additionally additionally ADV bracis-19071 8 2 , , PUNCT bracis-19071 8 3 certain certain ADJ bracis-19071 8 4 desirable desirable ADJ bracis-19071 8 5 features feature NOUN bracis-19071 8 6 are be AUX bracis-19071 8 7 not not PART bracis-19071 8 8 guaranteed guarantee VERB bracis-19071 8 9 , , PUNCT bracis-19071 8 10 such such ADJ bracis-19071 8 11 as as ADP bracis-19071 8 12 single single ADJ bracis-19071 8 13 , , PUNCT bracis-19071 8 14 similarly similarly ADV bracis-19071 8 15 - - PUNCT bracis-19071 8 16 sized sized ADJ bracis-19071 8 17 objects object NOUN bracis-19071 8 18 at at ADP bracis-19071 8 19 the the DET bracis-19071 8 20 center center NOUN bracis-19071 8 21 of of ADP bracis-19071 8 22 each each DET bracis-19071 8 23 picture picture NOUN bracis-19071 8 24 , , PUNCT bracis-19071 8 25 which which PRON bracis-19071 8 26 makes make VERB bracis-19071 8 27 the the DET bracis-19071 8 28 task task NOUN bracis-19071 8 29 more more ADV bracis-19071 8 30 challenging challenging ADJ bracis-19071 8 31 . . PUNCT bracis-19071 9 1 our our PRON bracis-19071 9 2 benchmark benchmark NOUN bracis-19071 9 3 results result NOUN bracis-19071 9 4 show show VERB bracis-19071 9 5 that that SCONJ bracis-19071 9 6 the the DET bracis-19071 9 7 dataset dataset NOUN bracis-19071 9 8 can can AUX bracis-19071 9 9 be be AUX bracis-19071 9 10 challenging challenge VERB bracis-19071 9 11 for for ADP bracis-19071 9 12 traditional traditional ADJ bracis-19071 9 13 machine machine NOUN bracis-19071 9 14 learning learn VERB bracis-19071 9 15 algorithms algorithm NOUN bracis-19071 9 16 without without ADP bracis-19071 9 17 any any DET bracis-19071 9 18 pre pre ADJ bracis-19071 9 19 - - ADJ bracis-19071 9 20 processing processing ADJ bracis-19071 9 21 steps step NOUN bracis-19071 9 22 , , PUNCT bracis-19071 9 23 while while SCONJ bracis-19071 9 24 state state NOUN bracis-19071 9 25 of of ADP bracis-19071 9 26 the the DET bracis-19071 9 27 art art NOUN bracis-19071 9 28 deep deep ADJ bracis-19071 9 29 learning learning NOUN bracis-19071 9 30 architectures architecture NOUN bracis-19071 9 31 achieve achieve VERB bracis-19071 9 32 around around ADV bracis-19071 9 33 82 82 NUM bracis-19071 9 34 % % NOUN bracis-19071 9 35 accuracy accuracy NOUN bracis-19071 9 36 , , PUNCT bracis-19071 9 37 which which PRON bracis-19071 9 38 means mean VERB bracis-19071 9 39 some some DET bracis-19071 9 40 effort effort NOUN bracis-19071 9 41 will will AUX bracis-19071 9 42 be be AUX bracis-19071 9 43 necessary necessary ADJ bracis-19071 9 44 to to PART bracis-19071 9 45 drive drive VERB bracis-19071 9 46 this this DET bracis-19071 9 47 accuracy accuracy NOUN bracis-19071 9 48 closer close ADV bracis-19071 9 49 to to ADP bracis-19071 9 50 what what PRON bracis-19071 9 51 has have AUX bracis-19071 9 52 been be AUX bracis-19071 9 53 accomplished accomplish VERB bracis-19071 9 54 for for ADP bracis-19071 9 55 mnist mnist NOUN bracis-19071 9 56 and and CCONJ bracis-19071 9 57 imagenet imagenet NOUN bracis-19071 9 58 . . PUNCT bracis-19071 10 1 access access NOUN bracis-19071 10 2 provided provide VERB bracis-19071 10 3 by by ADP bracis-19071 10 4 university university PROPN bracis-19071 10 5 of of ADP bracis-19071 10 6 notre notre PROPN bracis-19071 10 7 dame dame PROPN bracis-19071 10 8 hesburgh hesburgh PROPN bracis-19071 10 9 library library PROPN bracis-19071 10 10 . . PUNCT bracis-19071 11 1 download download PROPN bracis-19071 11 2 conference conference NOUN bracis-19071 11 3 paper paper NOUN bracis-19071 11 4 pdf pdf NOUN bracis-19071 11 5 similar similar ADJ bracis-19071 11 6 content content NOUN bracis-19071 11 7 being be AUX bracis-19071 11 8 viewed view VERB bracis-19071 11 9 by by ADP bracis-19071 11 10 others other NOUN bracis-19071 11 11 assessment assessment NOUN bracis-19071 11 12 of of ADP bracis-19071 11 13 iris iris NOUN bracis-19071 11 14 flower flower NOUN bracis-19071 11 15 classification classification NOUN bracis-19071 11 16 using use VERB bracis-19071 11 17 machine machine NOUN bracis-19071 11 18 learning learn VERB bracis-19071 11 19 algorithms algorithm NOUN bracis-19071 11 20 chapter chapter NOUN bracis-19071 11 21 © © PROPN bracis-19071 11 22 2021 2021 NUM bracis-19071 11 23 end end NOUN bracis-19071 11 24 - - PUNCT bracis-19071 11 25 to to ADP bracis-19071 11 26 - - PUNCT bracis-19071 11 27 end end NOUN bracis-19071 11 28 automated automate VERB bracis-19071 11 29 iris iris NOUN bracis-19071 11 30 segmentation segmentation NOUN bracis-19071 11 31 framework framework NOUN bracis-19071 11 32 using use VERB bracis-19071 11 33 u u NOUN bracis-19071 11 34 - - ADJ bracis-19071 11 35 net net ADJ bracis-19071 11 36 convolutional convolutional ADJ bracis-19071 11 37 neural neural ADJ bracis-19071 11 38 network network NOUN bracis-19071 11 39 chapter chapter NOUN bracis-19071 11 40 © © PROPN bracis-19071 11 41 2020 2020 NUM bracis-19071 11 42 flower flower NOUN bracis-19071 11 43 species species NOUN bracis-19071 11 44 detection detection NOUN bracis-19071 11 45 system system NOUN bracis-19071 11 46 using use VERB bracis-19071 11 47 deep deep ADJ bracis-19071 11 48 convolutional convolutional ADJ bracis-19071 11 49 neural neural ADJ bracis-19071 11 50 networks network NOUN bracis-19071 11 51 chapter chapter NOUN bracis-19071 11 52 © © PROPN bracis-19071 11 53 2022 2022 NUM bracis-19071 11 54 explore explore VERB bracis-19071 11 55 related relate VERB bracis-19071 11 56 subjects subject NOUN bracis-19071 11 57 discover discover VERB bracis-19071 11 58 the the DET bracis-19071 11 59 latest late ADJ bracis-19071 11 60 articles article NOUN bracis-19071 11 61 , , PUNCT bracis-19071 11 62 books book NOUN bracis-19071 11 63 and and CCONJ bracis-19071 11 64 news news NOUN bracis-19071 11 65 in in ADP bracis-19071 11 66 related related ADJ bracis-19071 11 67 subjects subject NOUN bracis-19071 11 68 , , PUNCT bracis-19071 11 69 suggested suggest VERB bracis-19071 11 70 using use VERB bracis-19071 11 71 machine machine NOUN bracis-19071 11 72 learning learning NOUN bracis-19071 11 73 . . PUNCT bracis-19071 12 1 categorization categorization NOUN bracis-19071 12 2 computer computer NOUN bracis-19071 12 3 vision vision PROPN bracis-19071 12 4 data data PROPN bracis-19071 12 5 science science PROPN bracis-19071 12 6 data data PROPN bracis-19071 12 7 mining mining NOUN bracis-19071 12 8 machine machine NOUN bracis-19071 12 9 learning learn VERB bracis-19071 12 10 object object NOUN bracis-19071 12 11 recognition recognition NOUN bracis-19071 12 12 1 1 NUM bracis-19071 12 13 introduction introduction NOUN bracis-19071 12 14 since since SCONJ bracis-19071 12 15 fisher fisher PROPN bracis-19071 12 16 ’s ’s PART bracis-19071 12 17 original original ADJ bracis-19071 12 18 publication publication NOUN bracis-19071 12 19 in in ADP bracis-19071 12 20 the the DET bracis-19071 12 21 context context NOUN bracis-19071 12 22 of of ADP bracis-19071 12 23 linear linear ADJ bracis-19071 12 24 discriminant discriminant ADJ bracis-19071 12 25 analysis analysis NOUN bracis-19071 12 26 [ [ X bracis-19071 12 27 4 4 NUM bracis-19071 12 28 ] ] PUNCT bracis-19071 12 29 , , PUNCT bracis-19071 12 30 the the DET bracis-19071 12 31 iris iris NOUN bracis-19071 12 32 flower flower NOUN bracis-19071 12 33 dataset dataset NOUN bracis-19071 12 34 turned turn VERB bracis-19071 12 35 into into ADP bracis-19071 12 36 one one NUM bracis-19071 12 37 of of ADP bracis-19071 12 38 the the DET bracis-19071 12 39 most most ADV bracis-19071 12 40 well well ADV bracis-19071 12 41 - - PUNCT bracis-19071 12 42 known know VERB bracis-19071 12 43 and and CCONJ bracis-19071 12 44 explored explore VERB bracis-19071 12 45 datasets dataset NOUN bracis-19071 12 46 in in ADP bracis-19071 12 47 statistical statistical ADJ bracis-19071 12 48 classification classification NOUN bracis-19071 12 49 and and CCONJ bracis-19071 12 50 machine machine NOUN bracis-19071 12 51 learning learning NOUN bracis-19071 12 52 ( ( PUNCT bracis-19071 12 53 ml ml NOUN bracis-19071 12 54 ) ) PUNCT bracis-19071 12 55 , , PUNCT bracis-19071 12 56 with with ADP bracis-19071 12 57 over over ADP bracis-19071 12 58 18,000 18,000 NUM bracis-19071 12 59 citations citation NOUN bracis-19071 12 60 until until ADP bracis-19071 12 61 the the DET bracis-19071 12 62 date date NOUN bracis-19071 12 63 of of ADP bracis-19071 12 64 publication publication NOUN bracis-19071 12 65 of of ADP bracis-19071 12 66 this this DET bracis-19071 12 67 work work NOUN bracis-19071 12 68 . . PUNCT bracis-19071 13 1 the the DET bracis-19071 13 2 dataset dataset NOUN bracis-19071 13 3 consists consist VERB bracis-19071 13 4 of of ADP bracis-19071 13 5 150 150 NUM bracis-19071 13 6 observations observation NOUN bracis-19071 13 7 , , PUNCT bracis-19071 13 8 equally equally ADV bracis-19071 13 9 divided divide VERB bracis-19071 13 10 into into ADP bracis-19071 13 11 three three NUM bracis-19071 13 12 classes class NOUN bracis-19071 13 13 ( ( PUNCT bracis-19071 13 14 iris iris NOUN bracis-19071 13 15 setosa setosa NOUN bracis-19071 13 16 , , PUNCT bracis-19071 13 17 iris iris NOUN bracis-19071 13 18 virginica virginica NOUN bracis-19071 13 19 and and CCONJ bracis-19071 13 20 iris iris NOUN bracis-19071 13 21 versicolor versicolor NOUN bracis-19071 13 22 ) ) PUNCT bracis-19071 13 23 , , PUNCT bracis-19071 13 24 which which PRON bracis-19071 13 25 are be AUX bracis-19071 13 26 described describe VERB bracis-19071 13 27 by by ADP bracis-19071 13 28 four four NUM bracis-19071 13 29 features feature NOUN bracis-19071 13 30 : : PUNCT bracis-19071 13 31 the the DET bracis-19071 13 32 length length NOUN bracis-19071 13 33 and and CCONJ bracis-19071 13 34 the the DET bracis-19071 13 35 width width NOUN bracis-19071 13 36 of of ADP bracis-19071 13 37 the the DET bracis-19071 13 38 sepals sepal NOUN bracis-19071 13 39 and and CCONJ bracis-19071 13 40 petals petal NOUN bracis-19071 13 41 , , PUNCT bracis-19071 13 42 in in ADP bracis-19071 13 43 centimeters centimeter NOUN bracis-19071 13 44 . . PUNCT bracis-19071 14 1 the the DET bracis-19071 14 2 measurements measurement NOUN bracis-19071 14 3 were be AUX bracis-19071 14 4 taken take VERB bracis-19071 14 5 by by ADP bracis-19071 14 6 edgar edgar PROPN bracis-19071 14 7 anderson anderson PROPN bracis-19071 15 1 [ [ X bracis-19071 15 2 1 1 NUM bracis-19071 15 3 ] ] PUNCT bracis-19071 15 4 , , PUNCT bracis-19071 15 5 who who PRON bracis-19071 15 6 was be AUX bracis-19071 15 7 interested interested ADJ bracis-19071 15 8 in in ADP bracis-19071 15 9 measuring measure VERB bracis-19071 15 10 the the DET bracis-19071 15 11 morphological morphological ADJ bracis-19071 15 12 variation variation NOUN bracis-19071 15 13 of of ADP bracis-19071 15 14 these these DET bracis-19071 15 15 species specie NOUN bracis-19071 15 16 , , PUNCT bracis-19071 15 17 while while SCONJ bracis-19071 15 18 fisher fisher PROPN bracis-19071 15 19 was be AUX bracis-19071 15 20 the the DET bracis-19071 15 21 first first ADJ bracis-19071 15 22 to to PART bracis-19071 15 23 use use VERB bracis-19071 15 24 the the DET bracis-19071 15 25 dataset dataset NOUN bracis-19071 15 26 in in ADP bracis-19071 15 27 a a DET bracis-19071 15 28 statistical statistical ADJ bracis-19071 15 29 learning learning NOUN bracis-19071 15 30 context context NOUN bracis-19071 15 31 . . PUNCT bracis-19071 16 1 iris iris NOUN bracis-19071 16 2 is be AUX bracis-19071 16 3 generally generally ADV bracis-19071 16 4 considered consider VERB bracis-19071 16 5 an an DET bracis-19071 16 6 easy easy ADJ bracis-19071 16 7 classification classification NOUN bracis-19071 16 8 problem problem NOUN bracis-19071 16 9 and and CCONJ bracis-19071 16 10 is be AUX bracis-19071 16 11 frequently frequently ADV bracis-19071 16 12 used use VERB bracis-19071 16 13 as as ADP bracis-19071 16 14 ml ml X bracis-19071 16 15 ’s ’s PART bracis-19071 16 16 “ " PUNCT bracis-19071 16 17 hello hello PROPN bracis-19071 16 18 world world NOUN bracis-19071 16 19 ” " PUNCT bracis-19071 16 20 ( ( PUNCT bracis-19071 16 21 i.e. i.e. X bracis-19071 16 22 , , PUNCT bracis-19071 16 23 as as ADP bracis-19071 16 24 the the DET bracis-19071 16 25 first first ADJ bracis-19071 16 26 example example NOUN bracis-19071 16 27 one one NUM bracis-19071 16 28 comes come VERB bracis-19071 16 29 into into ADP bracis-19071 16 30 contact contact NOUN bracis-19071 16 31 with with ADP bracis-19071 16 32 as as ADP bracis-19071 16 33 a a DET bracis-19071 16 34 beginner beginner NOUN bracis-19071 16 35 in in ADP bracis-19071 16 36 this this DET bracis-19071 16 37 area area NOUN bracis-19071 16 38 ) ) PUNCT bracis-19071 16 39 . . PUNCT bracis-19071 17 1 most most ADJ bracis-19071 17 2 classification classification NOUN bracis-19071 17 3 techniques technique NOUN bracis-19071 17 4 have have VERB bracis-19071 17 5 no no DET bracis-19071 17 6 trouble trouble NOUN bracis-19071 17 7 achieving achieve VERB bracis-19071 17 8 accuracies accuracy NOUN bracis-19071 17 9 well well ADV bracis-19071 17 10 above above ADP bracis-19071 17 11 90 90 NUM bracis-19071 17 12 % % NOUN bracis-19071 17 13 for for ADP bracis-19071 17 14 iris iris NOUN bracis-19071 17 15 with with ADP bracis-19071 17 16 various various ADJ bracis-19071 17 17 hyperparameter hyperparameter NOUN bracis-19071 17 18 configurations configuration NOUN bracis-19071 17 19 , , PUNCT bracis-19071 17 20 as as SCONJ bracis-19071 17 21 shown show VERB bracis-19071 17 22 by by ADP bracis-19071 17 23 multiple multiple ADJ bracis-19071 17 24 benchmark benchmark ADJ bracis-19071 17 25 results result NOUN bracis-19071 17 26 available available ADJ bracis-19071 17 27 at at ADP bracis-19071 17 28 openml openml NOUN bracis-19071 17 29 [ [ X bracis-19071 17 30 17 17 NUM bracis-19071 17 31 ] ] PUNCT bracis-19071 17 32 . . PUNCT bracis-19071 18 1 in in ADP bracis-19071 18 2 this this DET bracis-19071 18 3 paper paper NOUN bracis-19071 18 4 , , PUNCT bracis-19071 18 5 we we PRON bracis-19071 18 6 introduce introduce VERB bracis-19071 18 7 a a DET bracis-19071 18 8 new new ADJ bracis-19071 18 9 iris iris NOUN bracis-19071 18 10 flower flower NOUN bracis-19071 18 11 toy toy NOUN bracis-19071 18 12 dataset dataset NOUN bracis-19071 18 13 , , PUNCT bracis-19071 18 14 which which PRON bracis-19071 18 15 breaks break VERB bracis-19071 18 16 away away ADV bracis-19071 18 17 from from ADP bracis-19071 18 18 the the DET bracis-19071 18 19 original original ADJ bracis-19071 18 20 dataset dataset NOUN bracis-19071 18 21 ’s ’s PART bracis-19071 18 22 simplicity simplicity NOUN bracis-19071 18 23 by by ADP bracis-19071 18 24 turning turn VERB bracis-19071 18 25 the the DET bracis-19071 18 26 problem problem NOUN bracis-19071 18 27 of of ADP bracis-19071 18 28 iris iris NOUN bracis-19071 18 29 classification classification NOUN bracis-19071 18 30 into into ADP bracis-19071 18 31 a a DET bracis-19071 18 32 computer computer NOUN bracis-19071 18 33 vision vision NOUN bracis-19071 18 34 ( ( PUNCT bracis-19071 18 35 cv cv NOUN bracis-19071 18 36 ) ) PUNCT bracis-19071 18 37 task task NOUN bracis-19071 18 38 . . PUNCT bracis-19071 19 1 since since SCONJ bracis-19071 19 2 the the DET bracis-19071 19 3 rise rise NOUN bracis-19071 19 4 of of ADP bracis-19071 19 5 deep deep ADJ bracis-19071 19 6 learning learning NOUN bracis-19071 19 7 ( ( PUNCT bracis-19071 19 8 dl dl PROPN bracis-19071 19 9 ) ) PUNCT bracis-19071 19 10 , , PUNCT bracis-19071 19 11 the the DET bracis-19071 19 12 literature literature NOUN bracis-19071 19 13 of of ADP bracis-19071 19 14 cv cv PROPN bracis-19071 19 15 has have AUX bracis-19071 19 16 seen see VERB bracis-19071 19 17 many many ADJ bracis-19071 19 18 of of ADP bracis-19071 19 19 its its PRON bracis-19071 19 20 well well ADV bracis-19071 19 21 - - PUNCT bracis-19071 19 22 known know VERB bracis-19071 19 23 problems problem NOUN bracis-19071 19 24 , , PUNCT bracis-19071 19 25 including include VERB bracis-19071 19 26 mnist mnist NOUN bracis-19071 19 27 [ [ X bracis-19071 19 28 9 9 NUM bracis-19071 19 29 ] ] PUNCT bracis-19071 19 30 , , PUNCT bracis-19071 19 31 imagenet imagenet NOUN bracis-19071 20 1 [ [ X bracis-19071 20 2 3 3 NUM bracis-19071 20 3 ] ] PUNCT bracis-19071 20 4 , , PUNCT bracis-19071 20 5 and and CCONJ bracis-19071 20 6 the the DET bracis-19071 20 7 newer new ADJ bracis-19071 20 8 fashion fashion NOUN bracis-19071 20 9 - - PUNCT bracis-19071 20 10 mnist mnist NOUN bracis-19071 21 1 [ [ X bracis-19071 21 2 18 18 NUM bracis-19071 21 3 ] ] PUNCT bracis-19071 21 4 , , PUNCT bracis-19071 21 5 being be AUX bracis-19071 21 6 mostly mostly ADV bracis-19071 21 7 solved solve VERB bracis-19071 21 8 , , PUNCT bracis-19071 21 9 i.e. i.e. X bracis-19071 21 10 , , PUNCT bracis-19071 21 11 there there PRON bracis-19071 21 12 have have AUX bracis-19071 21 13 been be AUX bracis-19071 21 14 models model NOUN bracis-19071 21 15 which which PRON bracis-19071 21 16 achieved achieve VERB bracis-19071 21 17 near near ADP bracis-19071 21 18 - - PUNCT bracis-19071 21 19 human human ADJ bracis-19071 21 20 or or CCONJ bracis-19071 21 21 even even ADV bracis-19071 21 22 better well ADJ bracis-19071 21 23 - - PUNCT bracis-19071 21 24 than than ADP bracis-19071 21 25 - - PUNCT bracis-19071 21 26 human human ADJ bracis-19071 21 27 accuracy accuracy NOUN bracis-19071 21 28 for for ADP bracis-19071 21 29 these these DET bracis-19071 21 30 tasks task NOUN bracis-19071 21 31 . . PUNCT bracis-19071 22 1 hence hence ADV bracis-19071 22 2 , , PUNCT bracis-19071 22 3 this this DET bracis-19071 22 4 work work NOUN bracis-19071 22 5 introduces introduce VERB bracis-19071 22 6 a a DET bracis-19071 22 7 dataset dataset NOUN bracis-19071 22 8 which which PRON bracis-19071 22 9 features feature VERB bracis-19071 22 10 cv cv NOUN bracis-19071 22 11 challenges challenge NOUN bracis-19071 22 12 such such ADJ bracis-19071 22 13 as as ADP bracis-19071 22 14 fine fine ADJ bracis-19071 22 15 grained grain VERB bracis-19071 22 16 categorization categorization NOUN bracis-19071 22 17 , , PUNCT bracis-19071 22 18 background background NOUN bracis-19071 22 19 noise noise NOUN bracis-19071 22 20 , , PUNCT bracis-19071 22 21 real real ADJ bracis-19071 22 22 world world NOUN bracis-19071 22 23 environment environment NOUN bracis-19071 22 24 conditions condition NOUN bracis-19071 22 25 ( ( PUNCT bracis-19071 22 26 e.g. e.g. ADV bracis-19071 22 27 lighting lighting NOUN bracis-19071 22 28 variation variation NOUN bracis-19071 22 29 ) ) PUNCT bracis-19071 22 30 , , PUNCT bracis-19071 22 31 as as ADV bracis-19071 22 32 well well ADV bracis-19071 22 33 as as ADP bracis-19071 22 34 different different ADJ bracis-19071 22 35 scales scale NOUN bracis-19071 22 36 and and CCONJ bracis-19071 22 37 non non ADJ bracis-19071 22 38 - - ADJ bracis-19071 22 39 centered center VERB bracis-19071 22 40 objects object NOUN bracis-19071 22 41 . . PUNCT bracis-19071 23 1 the the DET bracis-19071 23 2 main main ADJ bracis-19071 23 3 idea idea NOUN bracis-19071 23 4 is be AUX bracis-19071 23 5 to to PART bracis-19071 23 6 make make VERB bracis-19071 23 7 this this DET bracis-19071 23 8 dataset dataset NOUN bracis-19071 23 9 available available ADJ bracis-19071 23 10 so so SCONJ bracis-19071 23 11 that that SCONJ bracis-19071 23 12 it it PRON bracis-19071 23 13 can can AUX bracis-19071 23 14 be be AUX bracis-19071 23 15 used use VERB bracis-19071 23 16 to to PART bracis-19071 23 17 validate validate VERB bracis-19071 23 18 machine machine NOUN bracis-19071 23 19 / / SYM bracis-19071 23 20 deep deep ADJ bracis-19071 23 21 learning learning NOUN bracis-19071 23 22 approaches approach NOUN bracis-19071 23 23 considering consider VERB bracis-19071 23 24 difficult difficult ADJ bracis-19071 23 25 cv cv ADJ bracis-19071 23 26 scenarios scenario NOUN bracis-19071 23 27 . . PUNCT bracis-19071 24 1 the the DET bracis-19071 24 2 paper paper NOUN bracis-19071 24 3 also also ADV bracis-19071 24 4 contains contain VERB bracis-19071 24 5 the the DET bracis-19071 24 6 classification classification NOUN bracis-19071 24 7 results result NOUN bracis-19071 24 8 of of ADP bracis-19071 24 9 traditional traditional ADJ bracis-19071 24 10 machine machine NOUN bracis-19071 24 11 learning learning NOUN bracis-19071 24 12 algorithms algorithm NOUN bracis-19071 24 13 , , PUNCT bracis-19071 24 14 as as ADV bracis-19071 24 15 well well ADV bracis-19071 24 16 as as ADP bracis-19071 24 17 some some DET bracis-19071 24 18 state state NOUN bracis-19071 24 19 - - PUNCT bracis-19071 24 20 of of ADP bracis-19071 24 21 - - PUNCT bracis-19071 24 22 the the DET bracis-19071 24 23 - - PUNCT bracis-19071 24 24 art art NOUN bracis-19071 24 25 deep deep ADJ bracis-19071 24 26 learning learning NOUN bracis-19071 24 27 architectures architecture NOUN bracis-19071 24 28 using use VERB bracis-19071 24 29 the the DET bracis-19071 24 30 iris iris NOUN bracis-19071 24 31 - - PUNCT bracis-19071 24 32 cv cv PROPN bracis-19071 24 33 dataset dataset NOUN bracis-19071 24 34 . . PUNCT bracis-19071 25 1 the the DET bracis-19071 25 2 remainder remainder NOUN bracis-19071 25 3 of of ADP bracis-19071 25 4 this this DET bracis-19071 25 5 paper paper NOUN bracis-19071 25 6 is be AUX bracis-19071 25 7 organized organize VERB bracis-19071 25 8 as as SCONJ bracis-19071 25 9 follows follow VERB bracis-19071 25 10 : : PUNCT bracis-19071 25 11 sect sect NOUN bracis-19071 25 12 . . PUNCT bracis-19071 25 13     SPACE bracis-19071 26 1 2 2 NUM bracis-19071 26 2 shows show VERB bracis-19071 26 3 some some DET bracis-19071 26 4 related relate VERB bracis-19071 26 5 works work NOUN bracis-19071 26 6 using use VERB bracis-19071 26 7 the the DET bracis-19071 26 8 inaturalist inaturalist ADJ bracis-19071 26 9 dataset dataset NOUN bracis-19071 26 10 , , PUNCT bracis-19071 26 11 sect sect NOUN bracis-19071 26 12 . . PUNCT bracis-19071 26 13     SPACE bracis-19071 27 1 3 3 NUM bracis-19071 27 2 describes describe VERB bracis-19071 27 3 our our PRON bracis-19071 27 4 new new ADJ bracis-19071 27 5 dataset dataset NOUN bracis-19071 27 6 , , PUNCT bracis-19071 27 7 sect sect NOUN bracis-19071 27 8 . . PUNCT bracis-19071 27 9     SPACE bracis-19071 28 1 4 4 NUM bracis-19071 28 2 presents present VERB bracis-19071 28 3 benchmark benchmark NOUN bracis-19071 28 4 results result NOUN bracis-19071 28 5 , , PUNCT bracis-19071 28 6 including include VERB bracis-19071 28 7 experiments experiment NOUN bracis-19071 28 8 with with ADP bracis-19071 28 9 traditional traditional ADJ bracis-19071 28 10 ml ml NOUN bracis-19071 28 11 algorithms algorithm NOUN bracis-19071 28 12 and and CCONJ bracis-19071 28 13 state state NOUN bracis-19071 28 14 of of ADP bracis-19071 28 15 the the DET bracis-19071 28 16 art art NOUN bracis-19071 28 17 deep deep ADJ bracis-19071 28 18 neural neural ADJ bracis-19071 28 19 networks network NOUN bracis-19071 28 20 , , PUNCT bracis-19071 28 21 and and CCONJ bracis-19071 28 22 finally finally ADV bracis-19071 28 23 , , PUNCT bracis-19071 28 24 sect sect NOUN bracis-19071 28 25 . . PUNCT bracis-19071 28 26     SPACE bracis-19071 28 27 5 5 NUM bracis-19071 28 28 contains contain VERB bracis-19071 28 29 our our PRON bracis-19071 28 30 final final ADJ bracis-19071 28 31 remarks remark NOUN bracis-19071 28 32 . . PUNCT bracis-19071 29 1 2 2 NUM bracis-19071 29 2 related relate VERB bracis-19071 29 3 works work NOUN bracis-19071 29 4 the the DET bracis-19071 29 5 inaturalist inaturalist NOUN bracis-19071 29 6 is be AUX bracis-19071 29 7 a a DET bracis-19071 29 8 dataset dataset NOUN bracis-19071 29 9 proposed propose VERB bracis-19071 29 10 by by ADP bracis-19071 29 11 [ [ X bracis-19071 29 12 16 16 NUM bracis-19071 29 13 ] ] PUNCT bracis-19071 29 14 and and CCONJ bracis-19071 29 15 has have VERB bracis-19071 29 16 675,170 675,170 NUM bracis-19071 29 17 images image NOUN bracis-19071 29 18 from from ADP bracis-19071 29 19 more more ADJ bracis-19071 29 20 than than ADP bracis-19071 29 21 5 5 NUM bracis-19071 29 22 thousand thousand NUM bracis-19071 29 23 different different ADJ bracis-19071 29 24 species specie NOUN bracis-19071 29 25 of of ADP bracis-19071 29 26 animals animal NOUN bracis-19071 29 27 and and CCONJ bracis-19071 29 28 plants plant NOUN bracis-19071 29 29 . . PUNCT bracis-19071 30 1 the the DET bracis-19071 30 2 category category NOUN bracis-19071 30 3 of of ADP bracis-19071 30 4 plants plant NOUN bracis-19071 30 5 has have VERB bracis-19071 30 6 almost almost ADV bracis-19071 30 7 200 200 NUM bracis-19071 30 8 thousand thousand NUM bracis-19071 30 9 images image NOUN bracis-19071 30 10 . . PUNCT bracis-19071 31 1 those those DET bracis-19071 31 2 species specie NOUN bracis-19071 31 3 images image NOUN bracis-19071 31 4 were be AUX bracis-19071 31 5 captured capture VERB bracis-19071 31 6 from from ADP bracis-19071 31 7 all all ADV bracis-19071 31 8 over over ADP bracis-19071 31 9 the the DET bracis-19071 31 10 world world NOUN bracis-19071 31 11 , , PUNCT bracis-19071 31 12 with with ADP bracis-19071 31 13 different different ADJ bracis-19071 31 14 cameras camera NOUN bracis-19071 31 15 and and CCONJ bracis-19071 31 16 image image NOUN bracis-19071 31 17 qualities quality NOUN bracis-19071 31 18 . . PUNCT bracis-19071 32 1 also also ADV bracis-19071 32 2 , , PUNCT bracis-19071 32 3 it it PRON bracis-19071 32 4 has have VERB bracis-19071 32 5 a a DET bracis-19071 32 6 large large ADJ bracis-19071 32 7 class class NOUN bracis-19071 32 8 imbalance imbalance NOUN bracis-19071 32 9 [ [ X bracis-19071 32 10 16 16 NUM bracis-19071 32 11 ] ] PUNCT bracis-19071 32 12 . . PUNCT bracis-19071 33 1 this this DET bracis-19071 33 2 dataset dataset NOUN bracis-19071 33 3 is be AUX bracis-19071 33 4 constantly constantly ADV bracis-19071 33 5 being be AUX bracis-19071 33 6 updated update VERB bracis-19071 33 7 . . PUNCT bracis-19071 34 1 although although SCONJ bracis-19071 34 2 inaturalist inaturalist PROPN bracis-19071 34 3 is be AUX bracis-19071 34 4 a a DET bracis-19071 34 5 toy toy NOUN bracis-19071 34 6 dataset dataset NOUN bracis-19071 34 7 that that PRON bracis-19071 34 8 was be AUX bracis-19071 34 9 used use VERB bracis-19071 34 10 in in ADP bracis-19071 34 11 kaggle kaggle ADJ bracis-19071 34 12 competitions competition NOUN bracis-19071 34 13 , , PUNCT bracis-19071 34 14 it it PRON bracis-19071 34 15 was be AUX bracis-19071 34 16 already already ADV bracis-19071 34 17 used use VERB bracis-19071 34 18 in in ADP bracis-19071 34 19 published publish VERB bracis-19071 34 20 papers paper NOUN bracis-19071 34 21 . . PUNCT bracis-19071 35 1 some some DET bracis-19071 35 2 works work NOUN bracis-19071 35 3 were be AUX bracis-19071 35 4 focused focus VERB bracis-19071 35 5 on on ADP bracis-19071 35 6 detection detection NOUN bracis-19071 35 7 [ [ X bracis-19071 35 8 16 16 NUM bracis-19071 35 9 ] ] PUNCT bracis-19071 35 10 and and CCONJ bracis-19071 35 11 others other NOUN bracis-19071 35 12 ’ ’ PART bracis-19071 35 13 purposes purpose NOUN bracis-19071 35 14 were be AUX bracis-19071 35 15 to to PART bracis-19071 35 16 make make VERB bracis-19071 35 17 classifications classification NOUN bracis-19071 35 18 such such ADJ bracis-19071 35 19 as as ADP bracis-19071 35 20 this this DET bracis-19071 35 21 work work NOUN bracis-19071 35 22 . . PUNCT bracis-19071 36 1 plant plant NOUN bracis-19071 36 2 classification classification NOUN bracis-19071 36 3 was be AUX bracis-19071 36 4 the the DET bracis-19071 36 5 focus focus NOUN bracis-19071 36 6 of of ADP bracis-19071 36 7 a a DET bracis-19071 36 8 study study NOUN bracis-19071 36 9 using use VERB bracis-19071 36 10 the the DET bracis-19071 36 11 inaturalist inaturalist ADJ bracis-19071 36 12 dataset dataset NOUN bracis-19071 37 1 [ [ X bracis-19071 37 2 12 12 NUM bracis-19071 37 3 ] ] PUNCT bracis-19071 37 4 , , PUNCT bracis-19071 37 5 where where SCONJ bracis-19071 37 6 the the DET bracis-19071 37 7 authors author NOUN bracis-19071 37 8 used use VERB bracis-19071 37 9 a a DET bracis-19071 37 10 convolutional convolutional ADJ bracis-19071 37 11 neural neural ADJ bracis-19071 37 12 network network NOUN bracis-19071 37 13 to to PART bracis-19071 37 14 classify classify VERB bracis-19071 37 15 the the DET bracis-19071 37 16 plant plant NOUN bracis-19071 37 17 ’s ’s PART bracis-19071 37 18 subclasses subclass NOUN bracis-19071 37 19 . . PUNCT bracis-19071 38 1 data datum NOUN bracis-19071 38 2 augmentation augmentation NOUN bracis-19071 38 3 was be AUX bracis-19071 38 4 used use VERB bracis-19071 38 5 to to PART bracis-19071 38 6 reduce reduce VERB bracis-19071 38 7 overfitting overfitting NOUN bracis-19071 38 8 and and CCONJ bracis-19071 38 9 balance balance VERB bracis-19071 38 10 the the DET bracis-19071 38 11 classes class NOUN bracis-19071 38 12 . . PUNCT bracis-19071 39 1 then then ADV bracis-19071 39 2 , , PUNCT bracis-19071 39 3 a a DET bracis-19071 39 4 transfer transfer NOUN bracis-19071 39 5 learning learn VERB bracis-19071 39 6 approach approach NOUN bracis-19071 39 7 based base VERB bracis-19071 39 8 on on ADP bracis-19071 39 9 resnet50 resnet50 NOUN bracis-19071 39 10 was be AUX bracis-19071 39 11 utilized utilize VERB bracis-19071 39 12 . . PUNCT bracis-19071 40 1 this this DET bracis-19071 40 2 work work NOUN bracis-19071 40 3 classifies classify VERB bracis-19071 40 4 different different ADJ bracis-19071 40 5 plant plant NOUN bracis-19071 40 6 species specie NOUN bracis-19071 40 7 rather rather ADV bracis-19071 40 8 than than ADP bracis-19071 40 9 only only ADJ bracis-19071 40 10 iris iris NOUN bracis-19071 40 11 unlike unlike ADP bracis-19071 40 12 our our PRON bracis-19071 40 13 work work NOUN bracis-19071 40 14 . . PUNCT bracis-19071 41 1 another another DET bracis-19071 41 2 study study NOUN bracis-19071 41 3 used use VERB bracis-19071 41 4 the the DET bracis-19071 41 5 entire entire ADJ bracis-19071 41 6 inaturalist inaturalist NOUN bracis-19071 41 7 dataset dataset NOUN bracis-19071 41 8 to to PART bracis-19071 41 9 make make VERB bracis-19071 41 10 both both CCONJ bracis-19071 41 11 classification classification NOUN bracis-19071 41 12 and and CCONJ bracis-19071 41 13 detection detection NOUN bracis-19071 41 14 [ [ X bracis-19071 41 15 16 16 NUM bracis-19071 41 16 ] ] PUNCT bracis-19071 41 17 . . PUNCT bracis-19071 42 1 for for ADP bracis-19071 42 2 classification classification NOUN bracis-19071 42 3 , , PUNCT bracis-19071 42 4 the the DET bracis-19071 42 5 following follow VERB bracis-19071 42 6 deep deep ADJ bracis-19071 42 7 network network NOUN bracis-19071 42 8 architectures architecture NOUN bracis-19071 42 9 were be AUX bracis-19071 42 10 performed perform VERB bracis-19071 42 11 : : PUNCT bracis-19071 42 12 resnets resnet NOUN bracis-19071 42 13 , , PUNCT bracis-19071 42 14 inception inception PROPN bracis-19071 42 15 v3 v3 PROPN bracis-19071 42 16 , , PUNCT bracis-19071 42 17 inception inception ADJ bracis-19071 42 18 resnetv2 resnetv2 NOUN bracis-19071 42 19 and and CCONJ bracis-19071 42 20 mobilenet mobilenet NOUN bracis-19071 42 21 . . PUNCT bracis-19071 43 1 from from ADP bracis-19071 43 2 those those DET bracis-19071 43 3 models model NOUN bracis-19071 43 4 , , PUNCT bracis-19071 43 5 inception inception ADJ bracis-19071 43 6 resnetv2 resnetv2 PROPN bracis-19071 43 7 se se PROPN bracis-19071 43 8 had have VERB bracis-19071 43 9 the the DET bracis-19071 43 10 best good ADJ bracis-19071 43 11 performance performance NOUN bracis-19071 43 12 . . PUNCT bracis-19071 44 1 this this DET bracis-19071 44 2 work work NOUN bracis-19071 44 3 also also ADV bracis-19071 44 4 classifies classify VERB bracis-19071 44 5 different different ADJ bracis-19071 44 6 plant plant NOUN bracis-19071 44 7 species specie NOUN bracis-19071 44 8 rather rather ADV bracis-19071 44 9 than than ADP bracis-19071 44 10 focusing focus VERB bracis-19071 44 11 on on ADP bracis-19071 44 12 iris iris NOUN bracis-19071 44 13 flowers flower NOUN bracis-19071 44 14 , , PUNCT bracis-19071 44 15 and and CCONJ bracis-19071 44 16 their their PRON bracis-19071 44 17 experiments experiment NOUN bracis-19071 44 18 did do AUX bracis-19071 44 19 not not PART bracis-19071 44 20 include include VERB bracis-19071 44 21 classic classic ADJ bracis-19071 44 22 algorithms algorithm NOUN bracis-19071 44 23 unlike unlike ADP bracis-19071 44 24 ours our NOUN bracis-19071 44 25 . . PUNCT bracis-19071 45 1 this this DET bracis-19071 45 2 work work NOUN bracis-19071 45 3 uses use VERB bracis-19071 45 4 the the DET bracis-19071 45 5 inaturalist inaturalist ADJ bracis-19071 45 6 dataset dataset NOUN bracis-19071 45 7 as as ADP bracis-19071 45 8 a a DET bracis-19071 45 9 single single ADJ bracis-19071 45 10 source source NOUN bracis-19071 45 11 , , PUNCT bracis-19071 45 12 but but CCONJ bracis-19071 45 13 some some DET bracis-19071 45 14 other other ADJ bracis-19071 45 15 researchers researcher NOUN bracis-19071 45 16 used use VERB bracis-19071 45 17 it it PRON bracis-19071 45 18 combined combine VERB bracis-19071 45 19 with with ADP bracis-19071 45 20 different different ADJ bracis-19071 45 21 datasets dataset NOUN bracis-19071 45 22 . . PUNCT bracis-19071 46 1 an an DET bracis-19071 46 2 example example NOUN bracis-19071 46 3 of of ADP bracis-19071 46 4 it it PRON bracis-19071 46 5 was be AUX bracis-19071 46 6 the the DET bracis-19071 46 7 work work NOUN bracis-19071 46 8 proposed propose VERB bracis-19071 46 9 by by ADP bracis-19071 46 10 [ [ X bracis-19071 46 11 7 7 NUM bracis-19071 46 12 ] ] PUNCT bracis-19071 46 13 . . PUNCT bracis-19071 47 1 after after ADP bracis-19071 47 2 selecting select VERB bracis-19071 47 3 different different ADJ bracis-19071 47 4 species specie NOUN bracis-19071 47 5 of of ADP bracis-19071 47 6 plant plant NOUN bracis-19071 47 7 images image NOUN bracis-19071 47 8 from from ADP bracis-19071 47 9 three three NUM bracis-19071 47 10 datasets dataset NOUN bracis-19071 47 11 , , PUNCT bracis-19071 47 12 they they PRON bracis-19071 47 13 applied apply VERB bracis-19071 47 14 deep deep ADJ bracis-19071 47 15 learning learning NOUN bracis-19071 47 16 techniques technique NOUN bracis-19071 47 17 for for ADP bracis-19071 47 18 plant plant NOUN bracis-19071 47 19 classification classification NOUN bracis-19071 47 20 . . PUNCT bracis-19071 48 1 their their PRON bracis-19071 48 2 goal goal NOUN bracis-19071 48 3 was be AUX bracis-19071 48 4 to to PART bracis-19071 48 5 achieve achieve VERB bracis-19071 48 6 at at ADP bracis-19071 48 7 least least ADV bracis-19071 48 8 50 50 NUM bracis-19071 48 9 percent percent NOUN bracis-19071 48 10 accuracy accuracy NOUN bracis-19071 48 11 as as ADP bracis-19071 48 12 a a DET bracis-19071 48 13 baseline baseline ADJ bracis-19071 48 14 classification classification NOUN bracis-19071 48 15 and and CCONJ bracis-19071 48 16 resnet50 resnet50 NOUN bracis-19071 48 17 was be AUX bracis-19071 48 18 able able ADJ bracis-19071 48 19 to to PART bracis-19071 48 20 classify classify VERB bracis-19071 48 21 almost almost ADV bracis-19071 48 22 half half NOUN bracis-19071 48 23 of of ADP bracis-19071 48 24 the the DET bracis-19071 48 25 inaturalist inaturalist ADJ bracis-19071 48 26 observations observation NOUN bracis-19071 48 27 . . PUNCT bracis-19071 49 1 the the DET bracis-19071 49 2 inaturalist inaturalist ADJ bracis-19071 49 3 dataset dataset NOUN bracis-19071 49 4 performed perform VERB bracis-19071 49 5 better well ADV bracis-19071 49 6 than than SCONJ bracis-19071 49 7 the the DET bracis-19071 49 8 portuguese portuguese ADJ bracis-19071 49 9 flora flora NOUN bracis-19071 49 10 dataset dataset VERB bracis-19071 49 11 , , PUNCT bracis-19071 49 12 but but CCONJ bracis-19071 49 13 the the DET bracis-19071 49 14 google google PROPN bracis-19071 49 15 image image NOUN bracis-19071 49 16 search search NOUN bracis-19071 49 17 observations observation NOUN bracis-19071 49 18 were be AUX bracis-19071 49 19 better well ADJ bracis-19071 49 20 than than ADP bracis-19071 49 21 both both PRON bracis-19071 49 22 . . PUNCT bracis-19071 50 1 as as SCONJ bracis-19071 50 2 the the DET bracis-19071 50 3 purpose purpose NOUN bracis-19071 50 4 of of ADP bracis-19071 50 5 this this DET bracis-19071 50 6 paper paper NOUN bracis-19071 50 7 is be AUX bracis-19071 50 8 to to PART bracis-19071 50 9 classify classify VERB bracis-19071 50 10 the the DET bracis-19071 50 11 different different ADJ bracis-19071 50 12 species specie NOUN bracis-19071 50 13 of of ADP bracis-19071 50 14 iris iris NOUN bracis-19071 50 15 , , PUNCT bracis-19071 50 16 the the DET bracis-19071 50 17 inaturalist inaturalist ADJ bracis-19071 50 18 dataset dataset NOUN bracis-19071 50 19 provided provide VERB bracis-19071 50 20 the the DET bracis-19071 50 21 images image NOUN bracis-19071 50 22 necessary necessary ADJ bracis-19071 50 23 to to PART bracis-19071 50 24 do do VERB bracis-19071 50 25 that that PRON bracis-19071 50 26 . . PUNCT bracis-19071 51 1 the the DET bracis-19071 51 2 paper paper NOUN bracis-19071 51 3 differs differ VERB bracis-19071 51 4 from from ADP bracis-19071 51 5 works work NOUN bracis-19071 51 6 described describe VERB bracis-19071 51 7 in in ADP bracis-19071 51 8 this this DET bracis-19071 51 9 section section NOUN bracis-19071 51 10 since since SCONJ bracis-19071 51 11 it it PRON bracis-19071 51 12 focuses focus VERB bracis-19071 51 13 on on ADP bracis-19071 51 14 benchmark benchmark ADJ bracis-19071 51 15 analysis analysis NOUN bracis-19071 51 16 , , PUNCT bracis-19071 51 17 thus thus ADV bracis-19071 51 18 it it PRON bracis-19071 51 19 covers cover VERB bracis-19071 51 20 a a DET bracis-19071 51 21 variety variety NOUN bracis-19071 51 22 of of ADP bracis-19071 51 23 algorithms algorithm NOUN bracis-19071 51 24 and and CCONJ bracis-19071 51 25 computational computational ADJ bracis-19071 51 26 cost cost NOUN bracis-19071 51 27 for for ADP bracis-19071 51 28 each each PRON bracis-19071 51 29 of of ADP bracis-19071 51 30 them they PRON bracis-19071 51 31 . . PUNCT bracis-19071 52 1 in in ADP bracis-19071 52 2 addition addition NOUN bracis-19071 52 3 , , PUNCT bracis-19071 52 4 we we PRON bracis-19071 52 5 focus focus VERB bracis-19071 52 6 specifically specifically ADV bracis-19071 52 7 on on ADP bracis-19071 52 8 iris iris NOUN bracis-19071 52 9 flowers flower NOUN bracis-19071 52 10 , , PUNCT bracis-19071 52 11 whereas whereas SCONJ bracis-19071 52 12 the the DET bracis-19071 52 13 other other ADJ bracis-19071 52 14 works work NOUN bracis-19071 52 15 used use VERB bracis-19071 52 16 plants plant NOUN bracis-19071 52 17 in in ADP bracis-19071 52 18 general general ADJ bracis-19071 52 19 . . PUNCT bracis-19071 53 1 3 3 NUM bracis-19071 53 2 a a DET bracis-19071 53 3 new new ADJ bracis-19071 53 4 iris iris NOUN bracis-19071 53 5 dataset dataset VERB bracis-19071 53 6 our our PRON bracis-19071 53 7 new new ADJ bracis-19071 53 8 iris iris NOUN bracis-19071 53 9 dataset dataset NOUN bracis-19071 53 10 , , PUNCT bracis-19071 53 11 called call VERB bracis-19071 53 12 iris iris NOUN bracis-19071 53 13 - - PUNCT bracis-19071 53 14 cv cv PROPN bracis-19071 53 15 , , PUNCT bracis-19071 53 16 consists consist VERB bracis-19071 53 17 of of ADP bracis-19071 53 18 5,139 5,139 NUM bracis-19071 53 19 examples example NOUN bracis-19071 53 20 extracting extract VERB bracis-19071 53 21 from from ADP bracis-19071 53 22 inaturalist inaturalist ADJ bracis-19071 53 23 ( ( PUNCT bracis-19071 53 24 september september PROPN bracis-19071 53 25 10th 10th NOUN bracis-19071 53 26 2020 2020 NUM bracis-19071 53 27 ) ) PUNCT bracis-19071 53 28 . . PUNCT bracis-19071 54 1 each each DET bracis-19071 54 2 example example NOUN bracis-19071 54 3 representing represent VERB bracis-19071 54 4 an an DET bracis-19071 54 5 rgb rgb PROPN bracis-19071 54 6 image image NOUN bracis-19071 54 7 associated associate VERB bracis-19071 54 8 with with ADP bracis-19071 54 9 a a DET bracis-19071 54 10 label label NOUN bracis-19071 54 11 corresponding correspond VERB bracis-19071 54 12 to to ADP bracis-19071 54 13 five five NUM bracis-19071 54 14 different different ADJ bracis-19071 54 15 species specie NOUN bracis-19071 54 16 : : PUNCT bracis-19071 54 17 bearded bearded ADJ bracis-19071 54 18 iris iris NOUN bracis-19071 54 19 ( ( PUNCT bracis-19071 54 20 be be AUX bracis-19071 54 21 ) ) PUNCT bracis-19071 54 22 ( ( PUNCT bracis-19071 54 23 iris iris NOUN bracis-19071 54 24 x x SYM bracis-19071 54 25 germanica germanica NOUN bracis-19071 54 26 , , PUNCT bracis-19071 54 27 928 928 NUM bracis-19071 54 28 images image NOUN bracis-19071 54 29 ) ) PUNCT bracis-19071 54 30 , , PUNCT bracis-19071 54 31 douglas douglas PROPN bracis-19071 54 32 iris iris PROPN bracis-19071 54 33 ( ( PUNCT bracis-19071 54 34 do do PROPN bracis-19071 54 35 ) ) PUNCT bracis-19071 54 36 ( ( PUNCT bracis-19071 54 37 iris iris NOUN bracis-19071 54 38 douglasiana douglasiana PROPN bracis-19071 54 39 , , PUNCT bracis-19071 54 40 944 944 NUM bracis-19071 54 41 images image NOUN bracis-19071 54 42 ) ) PUNCT bracis-19071 54 43 , , PUNCT bracis-19071 54 44 dwarf dwarf NOUN bracis-19071 54 45 crested crest VERB bracis-19071 54 46 iris iris NOUN bracis-19071 54 47 ( ( PUNCT bracis-19071 54 48 dw dw NOUN bracis-19071 54 49 ) ) PUNCT bracis-19071 54 50 ( ( PUNCT bracis-19071 54 51 iris iris NOUN bracis-19071 54 52 cristata cristata NOUN bracis-19071 54 53 , , PUNCT bracis-19071 54 54 1290 1290 NUM bracis-19071 54 55 images image NOUN bracis-19071 54 56 ) ) PUNCT bracis-19071 54 57 , , PUNCT bracis-19071 54 58 western western ADJ bracis-19071 54 59 blue blue ADJ bracis-19071 54 60 iris iris NOUN bracis-19071 54 61 ( ( PUNCT bracis-19071 54 62 we we PRON bracis-19071 54 63 ) ) PUNCT bracis-19071 54 64 ( ( PUNCT bracis-19071 54 65 iris iris NOUN bracis-19071 54 66 missouriensis missouriensis ADJ bracis-19071 54 67 , , PUNCT bracis-19071 54 68 1036 1036 NUM bracis-19071 54 69 images image NOUN bracis-19071 54 70 ) ) PUNCT bracis-19071 54 71 , , PUNCT bracis-19071 54 72 and and CCONJ bracis-19071 54 73 yellow yellow ADJ bracis-19071 54 74 iris iris NOUN bracis-19071 54 75 ( ( PUNCT bracis-19071 54 76 ye ye NOUN bracis-19071 54 77 ) ) PUNCT bracis-19071 54 78 ( ( PUNCT bracis-19071 54 79 iris iris NOUN bracis-19071 54 80 pseudacorus pseudacorus NOUN bracis-19071 54 81 , , PUNCT bracis-19071 54 82 941 941 NUM bracis-19071 54 83 images image NOUN bracis-19071 54 84 ) ) PUNCT bracis-19071 54 85 , , PUNCT bracis-19071 54 86 as as SCONJ bracis-19071 54 87 shown show VERB bracis-19071 54 88 in in ADP bracis-19071 54 89 table table NOUN bracis-19071 54 90     SPACE bracis-19071 54 91 1 1 NUM bracis-19071 54 92 . . PUNCT bracis-19071 55 1 all all DET bracis-19071 55 2 images image NOUN bracis-19071 55 3 were be AUX bracis-19071 55 4 gathered gather VERB bracis-19071 55 5 from from ADP bracis-19071 55 6 inaturalist inaturalist NOUN bracis-19071 55 7 [ [ X bracis-19071 55 8 10 10 NUM bracis-19071 55 9 ] ] PUNCT bracis-19071 55 10 , , PUNCT bracis-19071 55 11 a a DET bracis-19071 55 12 website website NOUN bracis-19071 55 13 that that PRON bracis-19071 55 14 provides provide VERB bracis-19071 55 15 creative creative ADJ bracis-19071 55 16 commons common NOUN bracis-19071 55 17 - - PUNCT bracis-19071 55 18 licensed license VERB bracis-19071 55 19 pictures picture NOUN bracis-19071 55 20 of of ADP bracis-19071 55 21 fauna fauna NOUN bracis-19071 55 22 and and CCONJ bracis-19071 55 23 flora flora NOUN bracis-19071 55 24 taken take VERB bracis-19071 55 25 by by ADP bracis-19071 55 26 users user NOUN bracis-19071 55 27 worldwide worldwide ADV bracis-19071 55 28 . . PUNCT bracis-19071 56 1 table table NOUN bracis-19071 56 2 1 1 NUM bracis-19071 56 3 . . PUNCT bracis-19071 56 4 dataset dataset ADJ bracis-19071 56 5 size size NOUN bracis-19071 56 6 per per ADP bracis-19071 56 7 classfull classfull NOUN bracis-19071 56 8 size size NOUN bracis-19071 56 9 table table NOUN bracis-19071 56 10 after after ADP bracis-19071 56 11 downloading download VERB bracis-19071 56 12 the the DET bracis-19071 56 13 images image NOUN bracis-19071 56 14 , , PUNCT bracis-19071 56 15 we we PRON bracis-19071 56 16 manually manually ADV bracis-19071 56 17 removed remove VERB bracis-19071 56 18 photos photo NOUN bracis-19071 56 19 that that PRON bracis-19071 56 20 had have VERB bracis-19071 56 21 too too ADV bracis-19071 56 22 much much ADJ bracis-19071 56 23 noise noise NOUN bracis-19071 56 24 , , PUNCT bracis-19071 56 25 i.e. i.e. X bracis-19071 56 26 , , PUNCT bracis-19071 56 27 pictures picture NOUN bracis-19071 56 28 of of ADP bracis-19071 56 29 many many ADJ bracis-19071 56 30 different different ADJ bracis-19071 56 31 flowers flower NOUN bracis-19071 56 32 , , PUNCT bracis-19071 56 33 human human ADJ bracis-19071 56 34 hands hand NOUN bracis-19071 56 35 covering cover VERB bracis-19071 56 36 the the DET bracis-19071 56 37 majority majority NOUN bracis-19071 56 38 of of ADP bracis-19071 56 39 the the DET bracis-19071 56 40 frame frame NOUN bracis-19071 56 41 , , PUNCT bracis-19071 56 42 and and CCONJ bracis-19071 56 43 blurry blurry ADJ bracis-19071 56 44 images image NOUN bracis-19071 56 45 , , PUNCT bracis-19071 56 46 which which PRON bracis-19071 56 47 could could AUX bracis-19071 56 48 confuse confuse VERB bracis-19071 56 49 learning learning NOUN bracis-19071 56 50 models model NOUN bracis-19071 56 51 , , PUNCT bracis-19071 56 52 resulting result VERB bracis-19071 56 53 in in ADP bracis-19071 56 54 a a DET bracis-19071 56 55 harder hard ADJ bracis-19071 56 56 problem problem NOUN bracis-19071 56 57 . . PUNCT bracis-19071 57 1 the the DET bracis-19071 57 2 original original ADJ bracis-19071 57 3 images image NOUN bracis-19071 57 4 have have VERB bracis-19071 57 5 different different ADJ bracis-19071 57 6 sizes size NOUN bracis-19071 57 7 , , PUNCT bracis-19071 57 8 therefore therefore ADV bracis-19071 57 9 we we PRON bracis-19071 57 10 resized resize VERB bracis-19071 57 11 them they PRON bracis-19071 57 12 to to ADP bracis-19071 57 13 a a DET bracis-19071 57 14 resolution resolution NOUN bracis-19071 57 15 of of ADP bracis-19071 57 16 256x256 256x256 PROPN bracis-19071 57 17 , , PUNCT bracis-19071 57 18 which which PRON bracis-19071 57 19 maintained maintain VERB bracis-19071 57 20 the the DET bracis-19071 57 21 images image NOUN bracis-19071 57 22 ’ ' PUNCT bracis-19071 57 23 main main ADJ bracis-19071 57 24 features feature NOUN bracis-19071 57 25 while while SCONJ bracis-19071 57 26 avoiding avoid VERB bracis-19071 57 27 high high ADJ bracis-19071 57 28 memory memory NOUN bracis-19071 57 29 requirements requirement NOUN bracis-19071 57 30 . . PUNCT bracis-19071 58 1 we we PRON bracis-19071 58 2 also also ADV bracis-19071 58 3 kept keep VERB bracis-19071 58 4 the the DET bracis-19071 58 5 color color NOUN bracis-19071 58 6 information information NOUN bracis-19071 58 7 , , PUNCT bracis-19071 58 8 which which PRON bracis-19071 58 9 is be AUX bracis-19071 58 10 coded code VERB bracis-19071 58 11 in in ADP bracis-19071 58 12 three three NUM bracis-19071 58 13 rgb rgb PROPN bracis-19071 58 14 channels channel NOUN bracis-19071 58 15 , , PUNCT bracis-19071 58 16 as as SCONJ bracis-19071 58 17 it it PRON bracis-19071 58 18 can can AUX bracis-19071 58 19 be be AUX bracis-19071 58 20 important important ADJ bracis-19071 58 21 to to PART bracis-19071 58 22 differentiate differentiate VERB bracis-19071 58 23 the the DET bracis-19071 58 24 classes class NOUN bracis-19071 58 25 . . PUNCT bracis-19071 59 1 table table NOUN bracis-19071 59 2     SPACE bracis-19071 59 3 2 2 NUM bracis-19071 59 4 shows show NOUN bracis-19071 59 5 resized resize VERB bracis-19071 59 6 examples example NOUN bracis-19071 59 7 of of ADP bracis-19071 59 8 each each DET bracis-19071 59 9 class class NOUN bracis-19071 59 10 . . PUNCT bracis-19071 60 1 the the DET bracis-19071 60 2 classes class NOUN bracis-19071 60 3 show show VERB bracis-19071 60 4 different different ADJ bracis-19071 60 5 color color NOUN bracis-19071 60 6 patterns pattern NOUN bracis-19071 60 7 , , PUNCT bracis-19071 60 8 particularly particularly ADV bracis-19071 60 9 the the DET bracis-19071 60 10 yellow yellow ADJ bracis-19071 60 11 iris iris NOUN bracis-19071 60 12 , , PUNCT bracis-19071 60 13 which which PRON bracis-19071 60 14 is be AUX bracis-19071 60 15 appropriately appropriately ADV bracis-19071 60 16 named name VERB bracis-19071 60 17 , , PUNCT bracis-19071 60 18 and and CCONJ bracis-19071 60 19 the the DET bracis-19071 60 20 dwarf dwarf NOUN bracis-19071 60 21 crested crest VERB bracis-19071 60 22 iris iris NOUN bracis-19071 60 23 , , PUNCT bracis-19071 60 24 with with ADP bracis-19071 60 25 its its PRON bracis-19071 60 26 white white ADJ bracis-19071 60 27 and and CCONJ bracis-19071 60 28 orange orange NOUN bracis-19071 60 29 crests crest NOUN bracis-19071 60 30 . . PUNCT bracis-19071 61 1 table table NOUN bracis-19071 61 2 2 2 NUM bracis-19071 61 3 . . PUNCT bracis-19071 61 4 class class NOUN bracis-19071 61 5 names name NOUN bracis-19071 61 6 and and CCONJ bracis-19071 61 7 examples example NOUN bracis-19071 61 8 from from ADP bracis-19071 61 9 the the DET bracis-19071 61 10 iris iris NOUN bracis-19071 61 11 - - PUNCT bracis-19071 61 12 cv cv PROPN bracis-19071 61 13 dataset.full dataset.full PROPN bracis-19071 61 14 size size NOUN bracis-19071 61 15 table table NOUN bracis-19071 61 16 the the DET bracis-19071 61 17 pictures picture NOUN bracis-19071 61 18 also also ADV bracis-19071 61 19 show show VERB bracis-19071 61 20 heavy heavy ADJ bracis-19071 61 21 background background NOUN bracis-19071 61 22 information information NOUN bracis-19071 61 23 , , PUNCT bracis-19071 61 24 thus thus ADV bracis-19071 61 25 models model NOUN bracis-19071 61 26 must must AUX bracis-19071 61 27 learn learn VERB bracis-19071 61 28 how how SCONJ bracis-19071 61 29 to to PART bracis-19071 61 30 separate separate VERB bracis-19071 61 31 the the DET bracis-19071 61 32 flowers flower NOUN bracis-19071 61 33 from from ADP bracis-19071 61 34 the the DET bracis-19071 61 35 background background NOUN bracis-19071 61 36 to to PART bracis-19071 61 37 perform perform VERB bracis-19071 61 38 well well ADV bracis-19071 61 39 . . PUNCT bracis-19071 62 1 images image NOUN bracis-19071 62 2 may may AUX bracis-19071 62 3 contain contain VERB bracis-19071 62 4 multiple multiple ADJ bracis-19071 62 5 and/or and/or CCONJ bracis-19071 62 6 non non ADJ bracis-19071 62 7 - - ADJ bracis-19071 62 8 centered centered ADJ bracis-19071 62 9 flowers flower NOUN bracis-19071 62 10 . . PUNCT bracis-19071 63 1 additionally additionally ADV bracis-19071 63 2 , , PUNCT bracis-19071 63 3 due due ADP bracis-19071 63 4 to to ADP bracis-19071 63 5 the the DET bracis-19071 63 6 different different ADJ bracis-19071 63 7 image image NOUN bracis-19071 63 8 and and CCONJ bracis-19071 63 9 flower flower NOUN bracis-19071 63 10 sizes size NOUN bracis-19071 63 11 , , PUNCT bracis-19071 63 12 some some DET bracis-19071 63 13 individuals individual NOUN bracis-19071 63 14 may may AUX bracis-19071 63 15 become become VERB bracis-19071 63 16 small small ADJ bracis-19071 63 17 compared compare VERB bracis-19071 63 18 to to ADP bracis-19071 63 19 others other NOUN bracis-19071 63 20 after after SCONJ bracis-19071 63 21 we we PRON bracis-19071 63 22 resize resize VERB bracis-19071 63 23 the the DET bracis-19071 63 24 pictures picture NOUN bracis-19071 63 25 , , PUNCT bracis-19071 63 26 as as SCONJ bracis-19071 63 27 seen see VERB bracis-19071 63 28 in in ADP bracis-19071 63 29 some some DET bracis-19071 63 30 examples example NOUN bracis-19071 63 31 in in ADP bracis-19071 63 32 table table NOUN bracis-19071 63 33     SPACE bracis-19071 63 34 2 2 NUM bracis-19071 63 35 . . PUNCT bracis-19071 64 1 there there PRON bracis-19071 64 2 can can AUX bracis-19071 64 3 also also ADV bracis-19071 64 4 be be AUX bracis-19071 64 5 very very ADV bracis-19071 64 6 different different ADJ bracis-19071 64 7 lighting lighting NOUN bracis-19071 64 8 conditions condition NOUN bracis-19071 64 9 across across ADP bracis-19071 64 10 pictures picture NOUN bracis-19071 64 11 . . PUNCT bracis-19071 65 1 finally finally ADV bracis-19071 65 2 , , PUNCT bracis-19071 65 3 successful successful ADJ bracis-19071 65 4 models model NOUN bracis-19071 65 5 will will AUX bracis-19071 65 6 have have VERB bracis-19071 65 7 to to PART bracis-19071 65 8 learn learn VERB bracis-19071 65 9 that that SCONJ bracis-19071 65 10 each each DET bracis-19071 65 11 species specie NOUN bracis-19071 65 12 can can AUX bracis-19071 65 13 show show VERB bracis-19071 65 14 different different ADJ bracis-19071 65 15 colors color NOUN bracis-19071 65 16 , , PUNCT bracis-19071 65 17 for for ADP bracis-19071 65 18 instance instance NOUN bracis-19071 65 19 , , PUNCT bracis-19071 65 20 dwarf dwarf PROPN bracis-19071 65 21 crested crest VERB bracis-19071 65 22 irises iris NOUN bracis-19071 65 23 can can AUX bracis-19071 65 24 be be AUX bracis-19071 65 25 lavender lavender NOUN bracis-19071 65 26 , , PUNCT bracis-19071 65 27 lilac lilac NOUN bracis-19071 65 28 , , PUNCT bracis-19071 65 29 pale pale ADJ bracis-19071 65 30 blue blue ADJ bracis-19071 65 31 , , PUNCT bracis-19071 65 32 purple purple ADJ bracis-19071 65 33 , , PUNCT bracis-19071 65 34 white white ADJ bracis-19071 65 35 , , PUNCT bracis-19071 65 36 or or CCONJ bracis-19071 65 37 pink pink ADJ bracis-19071 65 38 . . PUNCT bracis-19071 66 1 as as ADP bracis-19071 66 2 a a DET bracis-19071 66 3 result result NOUN bracis-19071 66 4 of of ADP bracis-19071 66 5 all all DET bracis-19071 66 6 these these DET bracis-19071 66 7 features feature NOUN bracis-19071 66 8 , , PUNCT bracis-19071 66 9 iris iris PROPN bracis-19071 66 10 - - PUNCT bracis-19071 66 11 cv cv PROPN bracis-19071 66 12 can can AUX bracis-19071 66 13 be be AUX bracis-19071 66 14 a a DET bracis-19071 66 15 challenging challenging ADJ bracis-19071 66 16 computer computer NOUN bracis-19071 66 17 vision vision NOUN bracis-19071 66 18 problem problem NOUN bracis-19071 66 19 . . PUNCT bracis-19071 67 1 4 4 NUM bracis-19071 67 2 benchmark benchmark NOUN bracis-19071 67 3 results result NOUN bracis-19071 67 4 we we PRON bracis-19071 67 5 begin begin VERB bracis-19071 67 6 our our PRON bracis-19071 67 7 experimental experimental ADJ bracis-19071 67 8 analysis analysis NOUN bracis-19071 67 9 by by ADP bracis-19071 67 10 validating validate VERB bracis-19071 67 11 our our PRON bracis-19071 67 12 dataset dataset NOUN bracis-19071 67 13 with with ADP bracis-19071 67 14 eight eight NUM bracis-19071 67 15 different different ADJ bracis-19071 67 16 classic classic ADJ bracis-19071 67 17 algorithms algorithm NOUN bracis-19071 67 18 using use VERB bracis-19071 67 19 scikit scikit NOUN bracis-19071 67 20 - - PUNCT bracis-19071 67 21 learn learn NOUN bracis-19071 67 22 and and CCONJ bracis-19071 67 23 xgboost xgboost NOUN bracis-19071 67 24 libraries library NOUN bracis-19071 67 25 , , PUNCT bracis-19071 67 26 as as SCONJ bracis-19071 67 27 listed list VERB bracis-19071 67 28 below below ADV bracis-19071 67 29 : : PUNCT bracis-19071 67 30 decision decision NOUN bracis-19071 67 31 tree tree NOUN bracis-19071 67 32 ( ( PUNCT bracis-19071 67 33 dt dt PROPN bracis-19071 67 34 ) ) PUNCT bracis-19071 67 35 ; ; PUNCT bracis-19071 67 36 extra extra ADJ bracis-19071 67 37 tree tree NOUN bracis-19071 67 38 ( ( PUNCT bracis-19071 67 39 et et NOUN bracis-19071 67 40 ) ) PUNCT bracis-19071 67 41 ; ; PUNCT bracis-19071 67 42 gradient gradient NOUN bracis-19071 67 43 boosting boosting NOUN bracis-19071 67 44 ( ( PUNCT bracis-19071 67 45 gb gb NOUN bracis-19071 67 46 ) ) PUNCT bracis-19071 67 47 ; ; PUNCT bracis-19071 67 48 extreme extreme ADJ bracis-19071 67 49 gradient gradient NOUN bracis-19071 67 50 boosting boost VERB bracis-19071 67 51 ( ( PUNCT bracis-19071 67 52 xgb xgb NUM bracis-19071 67 53 ) ) PUNCT bracis-19071 67 54 ; ; PUNCT bracis-19071 67 55 multilayer multilayer PROPN bracis-19071 67 56 perceptron perceptron PROPN bracis-19071 67 57 ( ( PUNCT bracis-19071 67 58 mlp mlp PROPN bracis-19071 67 59 ) ) PUNCT bracis-19071 67 60 ; ; PUNCT bracis-19071 67 61 perceptron perceptron PROPN bracis-19071 67 62 ; ; PUNCT bracis-19071 67 63 random random ADJ bracis-19071 67 64 forest forest NOUN bracis-19071 67 65 ( ( PUNCT bracis-19071 67 66 rf rf NOUN bracis-19071 67 67 ) ) PUNCT bracis-19071 67 68 ; ; PUNCT bracis-19071 67 69 k k X bracis-19071 67 70 - - PUNCT bracis-19071 67 71 nearest near ADJ bracis-19071 67 72 neighbors neighbor NOUN bracis-19071 67 73 ( ( PUNCT bracis-19071 67 74 knn knn PROPN bracis-19071 67 75 ) ) PUNCT bracis-19071 67 76 . . PUNCT bracis-19071 68 1 since since SCONJ bracis-19071 68 2 these these DET bracis-19071 68 3 algorithms algorithm NOUN bracis-19071 68 4 are be AUX bracis-19071 68 5 not not PART bracis-19071 68 6 originally originally ADV bracis-19071 68 7 equipped equip VERB bracis-19071 68 8 to to PART bracis-19071 68 9 receive receive VERB bracis-19071 68 10 pixel pixel NOUN bracis-19071 68 11 matrices matrix NOUN bracis-19071 68 12 as as ADP bracis-19071 68 13 input input NOUN bracis-19071 68 14 , , PUNCT bracis-19071 68 15 we we PRON bracis-19071 68 16 flattened flatten VERB bracis-19071 68 17 the the DET bracis-19071 68 18 256x256x3 256x256x3 NUM bracis-19071 68 19 images image NOUN bracis-19071 68 20 , , PUNCT bracis-19071 68 21 obtaining obtain VERB bracis-19071 68 22 vectors vector NOUN bracis-19071 68 23 with with ADP bracis-19071 68 24 196,608 196,608 NUM bracis-19071 68 25 dimensions dimension NOUN bracis-19071 68 26 which which PRON bracis-19071 68 27 can can AUX bracis-19071 68 28 then then ADV bracis-19071 68 29 be be AUX bracis-19071 68 30 used use VERB bracis-19071 68 31 as as ADP bracis-19071 68 32 input input NOUN bracis-19071 68 33 . . PUNCT bracis-19071 69 1 after after ADP bracis-19071 69 2 obtaining obtain VERB bracis-19071 69 3 the the DET bracis-19071 69 4 input input NOUN bracis-19071 69 5 vectors vector NOUN bracis-19071 69 6 , , PUNCT bracis-19071 69 7 we we PRON bracis-19071 69 8 rescaled rescale VERB bracis-19071 69 9 the the DET bracis-19071 69 10 images image NOUN bracis-19071 69 11 by by ADP bracis-19071 69 12 dividing divide VERB bracis-19071 69 13 each each DET bracis-19071 69 14 pixel pixel NOUN bracis-19071 69 15 by by ADP bracis-19071 69 16 255 255 NUM bracis-19071 69 17 before before ADP bracis-19071 69 18 training training NOUN bracis-19071 69 19 and and CCONJ bracis-19071 69 20 testing test VERB bracis-19071 69 21 the the DET bracis-19071 69 22 algorithms algorithm NOUN bracis-19071 69 23 . . PUNCT bracis-19071 70 1 experiments experiment NOUN bracis-19071 70 2 were be AUX bracis-19071 70 3 carried carry VERB bracis-19071 70 4 out out ADP bracis-19071 70 5 using use VERB bracis-19071 70 6 5 5 NUM bracis-19071 70 7 - - ADJ bracis-19071 70 8 fold fold ADJ bracis-19071 70 9 stratified stratified ADJ bracis-19071 70 10 cross cross NOUN bracis-19071 70 11 - - NOUN bracis-19071 70 12 validation validation NOUN bracis-19071 70 13 , , PUNCT bracis-19071 70 14 which which PRON bracis-19071 70 15 maintains maintain VERB bracis-19071 70 16 class class NOUN bracis-19071 70 17 proportions proportion NOUN bracis-19071 70 18 across across ADP bracis-19071 70 19 folds fold NOUN bracis-19071 70 20 . . PUNCT bracis-19071 71 1 the the DET bracis-19071 71 2 code code NOUN bracis-19071 71 3 was be AUX bracis-19071 71 4 implemented implement VERB bracis-19071 71 5 using use VERB bracis-19071 71 6 python python PROPN bracis-19071 71 7 ’s ’s PART bracis-19071 71 8 scikit scikit NOUN bracis-19071 71 9 - - PUNCT bracis-19071 71 10 learn learn VERB bracis-19071 71 11 library library NOUN bracis-19071 71 12 [ [ X bracis-19071 71 13 11 11 NUM bracis-19071 71 14 ] ] PUNCT bracis-19071 71 15 . . PUNCT bracis-19071 72 1 the the DET bracis-19071 72 2 algorithms algorithm NOUN bracis-19071 72 3 and and CCONJ bracis-19071 72 4 their their PRON bracis-19071 72 5 hyperparameter hyperparameter NOUN bracis-19071 72 6 values value NOUN bracis-19071 72 7 were be AUX bracis-19071 72 8 chosen choose VERB bracis-19071 72 9 based base VERB bracis-19071 72 10 on on ADP bracis-19071 72 11 their their PRON bracis-19071 72 12 performances performance NOUN bracis-19071 72 13 on on ADP bracis-19071 72 14 fashion fashion NOUN bracis-19071 72 15 - - PUNCT bracis-19071 72 16 mnist mnist NOUN bracis-19071 72 17 ’s ’s PART bracis-19071 72 18 benchmark benchmark NOUN bracis-19071 73 1 [ [ X bracis-19071 73 2 18 18 NUM bracis-19071 73 3 ] ] PUNCT bracis-19071 73 4 and and CCONJ bracis-19071 73 5 we we PRON bracis-19071 73 6 used use VERB bracis-19071 73 7 several several ADJ bracis-19071 73 8 machines machine NOUN bracis-19071 73 9 to to PART bracis-19071 73 10 streamline streamline VERB bracis-19071 73 11 the the DET bracis-19071 73 12 execution execution NOUN bracis-19071 73 13 of of ADP bracis-19071 73 14 tasks task NOUN bracis-19071 73 15 . . PUNCT bracis-19071 74 1 despite despite SCONJ bracis-19071 74 2 each each DET bracis-19071 74 3 machine machine NOUN bracis-19071 74 4 has have VERB bracis-19071 74 5 different different ADJ bracis-19071 74 6 settings setting NOUN bracis-19071 74 7 ( ( PUNCT bracis-19071 74 8 i.e. i.e. X bracis-19071 74 9 processor processor NOUN bracis-19071 74 10 , , PUNCT bracis-19071 74 11 graphic graphic ADJ bracis-19071 74 12 cards card NOUN bracis-19071 74 13 and and CCONJ bracis-19071 74 14 memory memory NOUN bracis-19071 74 15 ) ) PUNCT bracis-19071 74 16 , , PUNCT bracis-19071 74 17 it it PRON bracis-19071 74 18 is be AUX bracis-19071 74 19 important important ADJ bracis-19071 74 20 to to PART bracis-19071 74 21 notice notice VERB bracis-19071 74 22 that that SCONJ bracis-19071 74 23 this this PRON bracis-19071 74 24 only only ADV bracis-19071 74 25 affects affect VERB bracis-19071 74 26 the the DET bracis-19071 74 27 training training NOUN bracis-19071 74 28 time time NOUN bracis-19071 74 29 , , PUNCT bracis-19071 74 30 not not PART bracis-19071 74 31 the the DET bracis-19071 74 32 algorithms algorithm NOUN bracis-19071 74 33 overall overall ADJ bracis-19071 74 34 performance performance NOUN bracis-19071 74 35 . . PUNCT bracis-19071 75 1 the the DET bracis-19071 75 2 best good ADJ bracis-19071 75 3 results result NOUN bracis-19071 75 4 achieved achieve VERB bracis-19071 75 5 per per ADP bracis-19071 75 6 classifier classifier NOUN bracis-19071 75 7 can can AUX bracis-19071 75 8 be be AUX bracis-19071 75 9 seen see VERB bracis-19071 75 10 in in ADP bracis-19071 75 11 table table NOUN bracis-19071 75 12     SPACE bracis-19071 75 13 3 3 NUM bracis-19071 75 14 . . PUNCT bracis-19071 76 1 we we PRON bracis-19071 76 2 tested test VERB bracis-19071 76 3 different different ADJ bracis-19071 76 4 values value NOUN bracis-19071 76 5 ( ( PUNCT bracis-19071 76 6 table table NOUN bracis-19071 76 7     SPACE bracis-19071 76 8 3 3 NUM bracis-19071 76 9 ) ) PUNCT bracis-19071 76 10 for for ADP bracis-19071 76 11 the the DET bracis-19071 76 12 following follow VERB bracis-19071 76 13 hyperparameters hyperparameter NOUN bracis-19071 76 14 : : PUNCT bracis-19071 76 15 criterion criterion NOUN bracis-19071 76 16 ; ; PUNCT bracis-19071 76 17 objective objective NOUN bracis-19071 76 18 ; ; PUNCT bracis-19071 76 19 splitter splitter NOUN bracis-19071 76 20 ; ; PUNCT bracis-19071 76 21 max_depth max_depth NOUN bracis-19071 76 22 ; ; PUNCT bracis-19071 76 23 loss loss NOUN bracis-19071 76 24 ; ; PUNCT bracis-19071 76 25 n_estimators n_estimator NOUN bracis-19071 76 26 ; ; PUNCT bracis-19071 76 27 activation activation NOUN bracis-19071 76 28 ; ; PUNCT bracis-19071 76 29 hidden_layer_sizes hidden_layer_size NOUN bracis-19071 76 30 ; ; PUNCT bracis-19071 76 31 penalty penalty NOUN bracis-19071 76 32 ; ; PUNCT bracis-19071 76 33 n_neighbors n_neighbor NOUN bracis-19071 76 34 ; ; PUNCT bracis-19071 76 35 weights weight NOUN bracis-19071 76 36 ; ; PUNCT bracis-19071 76 37 and and CCONJ bracis-19071 76 38 p. p. NOUN bracis-19071 76 39 see see VERB bracis-19071 76 40 scikit scikit NOUN bracis-19071 76 41 - - PUNCT bracis-19071 76 42 learn learn VERB bracis-19071 76 43 ’s ’s PART bracis-19071 76 44 documentation documentation NOUN bracis-19071 77 1 [ [ X bracis-19071 77 2 11 11 NUM bracis-19071 77 3 ] ] PUNCT bracis-19071 77 4 for for ADP bracis-19071 77 5 more more ADJ bracis-19071 77 6 details detail NOUN bracis-19071 77 7 about about ADP bracis-19071 77 8 these these DET bracis-19071 77 9 parameters parameter NOUN bracis-19071 77 10 . . PUNCT bracis-19071 78 1 in in ADP bracis-19071 78 2 addition addition NOUN bracis-19071 78 3 , , PUNCT bracis-19071 78 4 table table NOUN bracis-19071 78 5     SPACE bracis-19071 78 6 4 4 NUM bracis-19071 78 7 describes describe VERB bracis-19071 78 8 the the DET bracis-19071 78 9 hardware hardware NOUN bracis-19071 78 10 resources resource NOUN bracis-19071 78 11 for for ADP bracis-19071 78 12 each each DET bracis-19071 78 13 classifier classifier NOUN bracis-19071 78 14 . . PUNCT bracis-19071 79 1 table table NOUN bracis-19071 79 2 3 3 NUM bracis-19071 79 3 . . PUNCT bracis-19071 79 4 results result NOUN bracis-19071 79 5 with with ADP bracis-19071 79 6 standard standard ADJ bracis-19071 79 7 deviation deviation NOUN bracis-19071 79 8 ( ( PUNCT bracis-19071 79 9 std std NOUN bracis-19071 79 10 ) ) PUNCT bracis-19071 79 11 of of ADP bracis-19071 79 12 classic classic ADJ bracis-19071 79 13 algorithms algorithm NOUN bracis-19071 79 14 for for ADP bracis-19071 79 15 the the DET bracis-19071 79 16 iris iris NOUN bracis-19071 79 17 - - PUNCT bracis-19071 79 18 cv cv PROPN bracis-19071 79 19 dataset dataset NOUN bracis-19071 79 20 using use VERB bracis-19071 79 21 5 5 NUM bracis-19071 79 22 - - ADJ bracis-19071 79 23 fold fold ADJ bracis-19071 79 24 cross cross NOUN bracis-19071 79 25 validation validation NOUN bracis-19071 79 26 . . PUNCT bracis-19071 80 1 the the DET bracis-19071 80 2 time time NOUN bracis-19071 80 3 column column NOUN bracis-19071 80 4 refers refer VERB bracis-19071 80 5 only only ADV bracis-19071 80 6 to to ADP bracis-19071 80 7 training training NOUN bracis-19071 80 8 time time NOUN bracis-19071 80 9 . . PUNCT bracis-19071 81 1 hyperparameter hyperparameter NOUN bracis-19071 81 2 names name NOUN bracis-19071 81 3 are be AUX bracis-19071 81 4 shown show VERB bracis-19071 81 5 as as SCONJ bracis-19071 81 6 they they PRON bracis-19071 81 7 appear appear VERB bracis-19071 81 8 in in ADP bracis-19071 81 9 scikit scikit NOUN bracis-19071 81 10 - - PUNCT bracis-19071 81 11 learn learn VERB bracis-19071 81 12 ’s ’s PART bracis-19071 81 13 documentation.full documentation.full ADJ bracis-19071 81 14 size size NOUN bracis-19071 81 15 table table NOUN bracis-19071 81 16 table table NOUN bracis-19071 81 17 4 4 NUM bracis-19071 81 18 . . PUNCT bracis-19071 81 19 hardware hardware NOUN bracis-19071 81 20 settings setting NOUN bracis-19071 81 21 for for ADP bracis-19071 81 22 training training NOUN bracis-19071 81 23 and and CCONJ bracis-19071 81 24 evaluate evaluate VERB bracis-19071 81 25 each each DET bracis-19071 81 26 classifierfull classifierfull ADJ bracis-19071 81 27 size size NOUN bracis-19071 81 28 table table NOUN bracis-19071 81 29 even even ADV bracis-19071 81 30 though though SCONJ bracis-19071 81 31 most most ADJ bracis-19071 81 32 results result NOUN bracis-19071 81 33 surpassed surpass VERB bracis-19071 81 34 the the DET bracis-19071 81 35 expected expect VERB bracis-19071 81 36 accuracy accuracy NOUN bracis-19071 81 37 of of ADP bracis-19071 81 38 a a DET bracis-19071 81 39 random random ADJ bracis-19071 81 40 classifier classifier NOUN bracis-19071 81 41 ( ( PUNCT bracis-19071 81 42 0.251 0.251 NUM bracis-19071 81 43 ) ) PUNCT bracis-19071 81 44 , , PUNCT bracis-19071 81 45 the the DET bracis-19071 81 46 overall overall ADJ bracis-19071 81 47 performance performance NOUN bracis-19071 81 48 shows show VERB bracis-19071 81 49 how how SCONJ bracis-19071 81 50 difficult difficult ADJ bracis-19071 81 51 this this DET bracis-19071 81 52 problem problem NOUN bracis-19071 81 53 is be AUX bracis-19071 81 54 for for ADP bracis-19071 81 55 classic classic ADJ bracis-19071 81 56 methods method NOUN bracis-19071 81 57 without without ADP bracis-19071 81 58 any any DET bracis-19071 81 59 extra extra ADJ bracis-19071 81 60 preprocessing preprocessing NOUN bracis-19071 81 61 specifically specifically ADV bracis-19071 81 62 designed design VERB bracis-19071 81 63 to to PART bracis-19071 81 64 improve improve VERB bracis-19071 81 65 their their PRON bracis-19071 81 66 results result NOUN bracis-19071 81 67 . . PUNCT bracis-19071 82 1 best good ADJ bracis-19071 82 2 results result NOUN bracis-19071 82 3 were be AUX bracis-19071 82 4 obtained obtain VERB bracis-19071 82 5 by by ADP bracis-19071 82 6 xgboost xgboost ADV bracis-19071 82 7 with with ADP bracis-19071 82 8 500 500 NUM bracis-19071 82 9 estimators estimator NOUN bracis-19071 82 10 , , PUNCT bracis-19071 82 11 multi multi ADJ bracis-19071 82 12 : : PUNCT bracis-19071 82 13 softmax softmax NOUN bracis-19071 82 14 as as ADP bracis-19071 82 15 the the DET bracis-19071 82 16 split split NOUN bracis-19071 82 17 objective objective NOUN bracis-19071 82 18 , , PUNCT bracis-19071 82 19 and and CCONJ bracis-19071 82 20 3 3 NUM bracis-19071 82 21 as as ADP bracis-19071 82 22 the the DET bracis-19071 82 23 max max PROPN bracis-19071 82 24 depth depth NOUN bracis-19071 82 25 , , PUNCT bracis-19071 82 26 with with ADP bracis-19071 82 27 a a DET bracis-19071 82 28 mean mean ADJ bracis-19071 82 29 accuracy accuracy NOUN bracis-19071 82 30 of of ADP bracis-19071 82 31 0.614 0.614 NUM bracis-19071 82 32 . . PUNCT bracis-19071 83 1 results result NOUN bracis-19071 83 2 were be AUX bracis-19071 83 3 also also ADV bracis-19071 83 4 poor poor ADJ bracis-19071 83 5 compared compare VERB bracis-19071 83 6 to to ADP bracis-19071 83 7 fashion fashion NOUN bracis-19071 83 8 - - PUNCT bracis-19071 83 9 mnist mnist NOUN bracis-19071 83 10 [ [ X bracis-19071 83 11 18 18 NUM bracis-19071 83 12 ] ] PUNCT bracis-19071 83 13 and and CCONJ bracis-19071 83 14 mnist mnist NOUN bracis-19071 84 1 [ [ X bracis-19071 84 2 9 9 NUM bracis-19071 84 3 ] ] PUNCT bracis-19071 84 4 , , PUNCT bracis-19071 84 5 where where SCONJ bracis-19071 84 6 the the DET bracis-19071 84 7 same same ADJ bracis-19071 84 8 algorithms algorithm NOUN bracis-19071 84 9 can can AUX bracis-19071 84 10 reach reach VERB bracis-19071 84 11 accuracy accuracy NOUN bracis-19071 84 12 values value NOUN bracis-19071 84 13 over over ADP bracis-19071 84 14 0.85 0.85 NUM bracis-19071 84 15 and and CCONJ bracis-19071 84 16 0.95 0.95 NUM bracis-19071 84 17 , , PUNCT bracis-19071 84 18 respectively respectively ADV bracis-19071 84 19 . . PUNCT bracis-19071 85 1 as as SCONJ bracis-19071 85 2 mentioned mention VERB bracis-19071 85 3 in in ADP bracis-19071 85 4 sect sect NOUN bracis-19071 85 5 . . PUNCT bracis-19071 85 6     SPACE bracis-19071 86 1 3 3 NUM bracis-19071 86 2 , , PUNCT bracis-19071 86 3 our our PRON bracis-19071 86 4 dataset dataset NOUN bracis-19071 86 5 ’s ’s PART bracis-19071 86 6 difficulty difficulty NOUN bracis-19071 86 7 can can AUX bracis-19071 86 8 be be AUX bracis-19071 86 9 explained explain VERB bracis-19071 86 10 by by ADP bracis-19071 86 11 some some DET bracis-19071 86 12 factors factor NOUN bracis-19071 86 13 , , PUNCT bracis-19071 86 14 such such ADJ bracis-19071 86 15 as as ADP bracis-19071 86 16 somewhat somewhat ADV bracis-19071 86 17 high high ADJ bracis-19071 86 18 - - PUNCT bracis-19071 86 19 dimensional dimensional ADJ bracis-19071 86 20 images image NOUN bracis-19071 86 21 , , PUNCT bracis-19071 86 22 rgb rgb PROPN bracis-19071 86 23 channels channel NOUN bracis-19071 86 24 , , PUNCT bracis-19071 86 25 non non ADJ bracis-19071 86 26 - - ADJ bracis-19071 86 27 centered center VERB bracis-19071 86 28 images image NOUN bracis-19071 86 29 , , PUNCT bracis-19071 86 30 noisy noisy ADJ bracis-19071 86 31 backgrounds background NOUN bracis-19071 86 32 , , PUNCT bracis-19071 86 33 and and CCONJ bracis-19071 86 34 petals petal NOUN bracis-19071 86 35 with with ADP bracis-19071 86 36 different different ADJ bracis-19071 86 37 colors color NOUN bracis-19071 86 38 within within ADP bracis-19071 86 39 each each DET bracis-19071 86 40 species specie NOUN bracis-19071 86 41 . . PUNCT bracis-19071 87 1 thus thus ADV bracis-19071 87 2 , , PUNCT bracis-19071 87 3 this this DET bracis-19071 87 4 problem problem NOUN bracis-19071 87 5 can can AUX bracis-19071 87 6 not not PART bracis-19071 87 7 be be AUX bracis-19071 87 8 tackled tackle VERB bracis-19071 87 9 by by ADP bracis-19071 87 10 simple simple ADJ bracis-19071 87 11 approaches approach NOUN bracis-19071 87 12 . . PUNCT bracis-19071 88 1 4.1 4.1 NUM bracis-19071 88 2 deep deep ADJ bracis-19071 88 3 neural neural ADJ bracis-19071 88 4 net net ADJ bracis-19071 88 5 results result NOUN bracis-19071 88 6 we we PRON bracis-19071 88 7 now now ADV bracis-19071 88 8 turn turn VERB bracis-19071 88 9 to to ADP bracis-19071 88 10 state state NOUN bracis-19071 88 11 - - PUNCT bracis-19071 88 12 of of ADP bracis-19071 88 13 - - PUNCT bracis-19071 88 14 the the DET bracis-19071 88 15 - - PUNCT bracis-19071 88 16 art art NOUN bracis-19071 88 17 convolutional convolutional ADJ bracis-19071 88 18 neural neural ADJ bracis-19071 88 19 network network NOUN bracis-19071 88 20 ( ( PUNCT bracis-19071 88 21 cnn cnn PROPN bracis-19071 88 22 ) ) PUNCT bracis-19071 88 23 architectures architecture NOUN bracis-19071 88 24 , , PUNCT bracis-19071 88 25 implemented implement VERB bracis-19071 88 26 using use VERB bracis-19071 88 27 tensorflow tensorflow NOUN bracis-19071 88 28 2 2 NUM bracis-19071 88 29 [ [ X bracis-19071 88 30 5 5 NUM bracis-19071 88 31 ] ] PUNCT bracis-19071 88 32 , , PUNCT bracis-19071 88 33 to to PART bracis-19071 88 34 see see VERB bracis-19071 88 35 how how SCONJ bracis-19071 88 36 these these DET bracis-19071 88 37 techniques technique NOUN bracis-19071 88 38 performing perform VERB bracis-19071 88 39 in in ADP bracis-19071 88 40 the the DET bracis-19071 88 41 classification classification NOUN bracis-19071 88 42 of of ADP bracis-19071 88 43 iris iris NOUN bracis-19071 88 44 - - PUNCT bracis-19071 88 45 cv cv NOUN bracis-19071 88 46 images image NOUN bracis-19071 88 47 . . PUNCT bracis-19071 89 1 we we PRON bracis-19071 89 2 chose choose VERB bracis-19071 89 3 to to ADP bracis-19071 89 4 the the DET bracis-19071 89 5 hyperparameters hyperparameter NOUN bracis-19071 89 6 used use VERB bracis-19071 89 7 in in ADP bracis-19071 89 8 the the DET bracis-19071 89 9 densenet densenet NOUN bracis-19071 89 10 [ [ X bracis-19071 89 11 8 8 NUM bracis-19071 89 12 ] ] X bracis-19071 89 13 paper paper NOUN bracis-19071 89 14 , , PUNCT bracis-19071 89 15 since since SCONJ bracis-19071 89 16 this this PRON bracis-19071 89 17 is be AUX bracis-19071 89 18 a a DET bracis-19071 89 19 state state NOUN bracis-19071 89 20 - - PUNCT bracis-19071 89 21 of of ADP bracis-19071 89 22 - - PUNCT bracis-19071 89 23 the the DET bracis-19071 89 24 - - PUNCT bracis-19071 89 25 art art NOUN bracis-19071 89 26 network network NOUN bracis-19071 89 27 . . PUNCT bracis-19071 90 1 we we PRON bracis-19071 90 2 trained train VERB bracis-19071 90 3 each each DET bracis-19071 90 4 network network NOUN bracis-19071 90 5 with with ADP bracis-19071 90 6 the the DET bracis-19071 90 7 stochastic stochastic ADJ bracis-19071 90 8 gradient gradient ADJ bracis-19071 90 9 descent descent NOUN bracis-19071 90 10 ( ( PUNCT bracis-19071 90 11 sgd sgd NOUN bracis-19071 90 12 ) ) PUNCT bracis-19071 90 13 optimizer optimizer NOUN bracis-19071 90 14 , , PUNCT bracis-19071 90 15 using use VERB bracis-19071 90 16 0.9 0.9 NUM bracis-19071 90 17 for for ADP bracis-19071 90 18 the the DET bracis-19071 90 19 nesterov nesterov ADJ bracis-19071 90 20 momentum momentum NOUN bracis-19071 90 21 and and CCONJ bracis-19071 90 22 40 40 NUM bracis-19071 90 23 epochs epoch NOUN bracis-19071 90 24 . . PUNCT bracis-19071 91 1 regarding regard VERB bracis-19071 91 2 the the DET bracis-19071 91 3 learning learning NOUN bracis-19071 91 4 rate rate NOUN bracis-19071 91 5 , , PUNCT bracis-19071 91 6 we we PRON bracis-19071 91 7 chose choose VERB bracis-19071 91 8 a a DET bracis-19071 91 9 starting starting NOUN bracis-19071 91 10 value value NOUN bracis-19071 91 11 of of ADP bracis-19071 91 12 0.1 0.1 NUM bracis-19071 91 13 , , PUNCT bracis-19071 91 14 decaying decay VERB bracis-19071 91 15 at at ADP bracis-19071 91 16 a a DET bracis-19071 91 17 pace pace NOUN bracis-19071 91 18 of of ADP bracis-19071 91 19 \(10^\frac{epoch}{20}\ \(10^\frac{epoch}{20}\ NOUN bracis-19071 91 20 ) ) PUNCT bracis-19071 91 21 as as ADP bracis-19071 91 22 a a DET bracis-19071 91 23 way way NOUN bracis-19071 91 24 of of ADP bracis-19071 91 25 tuning tune VERB bracis-19071 91 26 network network NOUN bracis-19071 91 27 performance performance NOUN bracis-19071 91 28 , , PUNCT bracis-19071 91 29 which which PRON bracis-19071 91 30 was be AUX bracis-19071 91 31 observed observe VERB bracis-19071 91 32 empirically empirically ADV bracis-19071 91 33 . . PUNCT bracis-19071 92 1 in in ADP bracis-19071 92 2 addition addition NOUN bracis-19071 92 3 , , PUNCT bracis-19071 92 4 we we PRON bracis-19071 92 5 used use VERB bracis-19071 92 6 different different ADJ bracis-19071 92 7 state state NOUN bracis-19071 92 8 - - PUNCT bracis-19071 92 9 of of ADP bracis-19071 92 10 - - PUNCT bracis-19071 92 11 the the DET bracis-19071 92 12 - - PUNCT bracis-19071 92 13 art art NOUN bracis-19071 92 14 architectures architecture NOUN bracis-19071 92 15 such such ADJ bracis-19071 92 16 as as ADP bracis-19071 92 17 efficientnet efficientnet NOUN bracis-19071 92 18 , , PUNCT bracis-19071 92 19 mobilenet mobilenet NOUN bracis-19071 92 20 and and CCONJ bracis-19071 92 21 resnet50 resnet50 NOUN bracis-19071 92 22 to to PART bracis-19071 92 23 verify verify VERB bracis-19071 92 24 if if SCONJ bracis-19071 92 25 it it PRON bracis-19071 92 26 has have VERB bracis-19071 92 27 any any DET bracis-19071 92 28 performance performance NOUN bracis-19071 92 29 improvement improvement NOUN bracis-19071 92 30 regarding regard VERB bracis-19071 92 31 state state NOUN bracis-19071 92 32 - - PUNCT bracis-19071 92 33 of of ADP bracis-19071 92 34 - - PUNCT bracis-19071 92 35 the the DET bracis-19071 92 36 - - PUNCT bracis-19071 92 37 art art NOUN bracis-19071 92 38 network network NOUN bracis-19071 92 39 architectures architecture NOUN bracis-19071 92 40 . . PUNCT bracis-19071 93 1 also also ADV bracis-19071 93 2 , , PUNCT bracis-19071 93 3 we we PRON bracis-19071 93 4 performed perform VERB bracis-19071 93 5 data datum NOUN bracis-19071 93 6 augmentation augmentation NOUN bracis-19071 93 7 to to PART bracis-19071 93 8 balance balance VERB bracis-19071 93 9 the the DET bracis-19071 93 10 class class NOUN bracis-19071 93 11 proportions proportion NOUN bracis-19071 93 12 . . PUNCT bracis-19071 94 1 we we PRON bracis-19071 94 2 tested test VERB bracis-19071 94 3 several several ADJ bracis-19071 94 4 combinations combination NOUN bracis-19071 94 5 of of ADP bracis-19071 94 6 the the DET bracis-19071 94 7 parameters parameter NOUN bracis-19071 94 8 of of ADP bracis-19071 94 9 the the DET bracis-19071 94 10 imagedatagenerator imagedatagenerator NOUN bracis-19071 94 11 provided provide VERB bracis-19071 94 12 by by ADP bracis-19071 94 13 tensorflow tensorflow NOUN bracis-19071 94 14 and and CCONJ bracis-19071 94 15 keras keras PROPN bracis-19071 94 16 , , PUNCT bracis-19071 94 17 viewing view VERB bracis-19071 94 18 the the DET bracis-19071 94 19 images image NOUN bracis-19071 94 20 we we PRON bracis-19071 94 21 had have VERB bracis-19071 94 22 as as ADP bracis-19071 94 23 a a DET bracis-19071 94 24 result result NOUN bracis-19071 94 25 and and CCONJ bracis-19071 94 26 choosing choose VERB bracis-19071 94 27 the the DET bracis-19071 94 28 ones one NOUN bracis-19071 94 29 that that PRON bracis-19071 94 30 kept keep VERB bracis-19071 94 31 most most ADJ bracis-19071 94 32 of of ADP bracis-19071 94 33 the the DET bracis-19071 94 34 original original ADJ bracis-19071 94 35 information information NOUN bracis-19071 94 36 . . PUNCT bracis-19071 95 1 below below ADP bracis-19071 95 2 we we PRON bracis-19071 95 3 list list VERB bracis-19071 95 4 the the DET bracis-19071 95 5 augmentation augmentation NOUN bracis-19071 95 6 parameters parameter NOUN bracis-19071 95 7 : : PUNCT bracis-19071 95 8 rescale rescale NOUN bracis-19071 95 9 : : PUNCT bracis-19071 95 10 1/255 1/255 NUM bracis-19071 95 11 rotation rotation NOUN bracis-19071 95 12 range range NOUN bracis-19071 95 13 : : PUNCT bracis-19071 95 14 20\(^{\circ 20\(^{\circ NUM bracis-19071 95 15 } } PUNCT bracis-19071 96 1 \ \ NOUN bracis-19071 96 2 ) ) PUNCT bracis-19071 96 3 width width ADJ bracis-19071 96 4 shift shift NOUN bracis-19071 96 5 range range NOUN bracis-19071 96 6 : : PUNCT bracis-19071 96 7 0.1 0.1 NUM bracis-19071 96 8 height height NOUN bracis-19071 96 9 shift shift NOUN bracis-19071 96 10 range range NOUN bracis-19071 96 11 : : PUNCT bracis-19071 96 12 0.1 0.1 NUM bracis-19071 96 13 horizontal horizontal ADJ bracis-19071 96 14 flip flip NOUN bracis-19071 96 15 : : PUNCT bracis-19071 96 16 true true ADJ bracis-19071 96 17 shear shear NOUN bracis-19071 96 18 range range NOUN bracis-19071 96 19 : : PUNCT bracis-19071 96 20 0.1 0.1 NUM bracis-19071 96 21 zoom zoom NOUN bracis-19071 96 22 range range NOUN bracis-19071 96 23 : : PUNCT bracis-19071 96 24 0.4 0.4 NUM bracis-19071 96 25 0.5 0.5 NUM bracis-19071 96 26 fill fill NOUN bracis-19071 96 27 mode mode NOUN bracis-19071 96 28 : : PUNCT bracis-19071 96 29 nearest near ADJ bracis-19071 96 30 more more ADJ bracis-19071 96 31 information information NOUN bracis-19071 96 32 about about ADP bracis-19071 96 33 these these DET bracis-19071 96 34 settings setting NOUN bracis-19071 96 35 and and CCONJ bracis-19071 96 36 the the DET bracis-19071 96 37 process process NOUN bracis-19071 96 38 used use VERB bracis-19071 96 39 by by ADP bracis-19071 96 40 imagedatagenerator imagedatagenerator NOUN bracis-19071 96 41 is be AUX bracis-19071 96 42 available available ADJ bracis-19071 96 43 in in ADP bracis-19071 96 44 the the DET bracis-19071 96 45 official official ADJ bracis-19071 96 46 tensorflow tensorflow NOUN bracis-19071 96 47 documentationfootnote documentationfootnote NOUN bracis-19071 96 48 1 1 X bracis-19071 96 49 . . PUNCT bracis-19071 97 1 the the DET bracis-19071 97 2 trained train VERB bracis-19071 97 3 architectures architecture NOUN bracis-19071 97 4 are be AUX bracis-19071 97 5 listed list VERB bracis-19071 97 6 below below ADP bracis-19071 97 7 : : PUNCT bracis-19071 98 1 densenet121 densenet121 PROPN bracis-19071 98 2 . . PROPN bracis-19071 98 3 proposed propose VERB bracis-19071 98 4 by by ADP bracis-19071 98 5 huang huang PROPN bracis-19071 99 1 [ [ X bracis-19071 99 2 8 8 NUM bracis-19071 99 3 ] ] PUNCT bracis-19071 99 4 , , PUNCT bracis-19071 99 5 this this DET bracis-19071 99 6 deep deep ADJ bracis-19071 99 7 net net NOUN bracis-19071 99 8 aims aim VERB bracis-19071 99 9 to to PART bracis-19071 99 10 optimize optimize VERB bracis-19071 99 11 the the DET bracis-19071 99 12 flow flow NOUN bracis-19071 99 13 of of ADP bracis-19071 99 14 information information NOUN bracis-19071 99 15 between between ADP bracis-19071 99 16 the the DET bracis-19071 99 17 layers layer NOUN bracis-19071 99 18 of of ADP bracis-19071 99 19 the the DET bracis-19071 99 20 network network NOUN bracis-19071 99 21 , , PUNCT bracis-19071 99 22 making make VERB bracis-19071 99 23 a a DET bracis-19071 99 24 dependency dependency NOUN bracis-19071 99 25 link link NOUN bracis-19071 99 26 between between ADP bracis-19071 99 27 them they PRON bracis-19071 99 28 , , PUNCT bracis-19071 99 29 trying try VERB bracis-19071 99 30 to to PART bracis-19071 99 31 minimize minimize VERB bracis-19071 99 32 convergence convergence NOUN bracis-19071 99 33 time time NOUN bracis-19071 99 34 through through ADP bracis-19071 99 35 shorter short ADJ bracis-19071 99 36 paths path NOUN bracis-19071 99 37 . . PUNCT bracis-19071 100 1 efficientnetb0 efficientnetb0 PROPN bracis-19071 100 2 . . PUNCT bracis-19071 101 1 this this DET bracis-19071 101 2 architecture architecture NOUN bracis-19071 101 3 , , PUNCT bracis-19071 101 4 introduced introduce VERB bracis-19071 101 5 by by ADP bracis-19071 101 6 mingxing mingxe VERB bracis-19071 101 7 tan tan PROPN bracis-19071 101 8 [ [ X bracis-19071 101 9 15 15 NUM bracis-19071 101 10 ] ] PUNCT bracis-19071 101 11 , , PUNCT bracis-19071 101 12 is be AUX bracis-19071 101 13 a a DET bracis-19071 101 14 neural neural ADJ bracis-19071 101 15 network network NOUN bracis-19071 101 16 obtained obtain VERB bracis-19071 101 17 through through ADP bracis-19071 101 18 a a DET bracis-19071 101 19 compound compound NOUN bracis-19071 101 20 scaling scale VERB bracis-19071 101 21 method method NOUN bracis-19071 101 22 of of ADP bracis-19071 101 23 convnets convnet NOUN bracis-19071 101 24 ’s ’s PART bracis-19071 101 25 depth depth NOUN bracis-19071 101 26 , , PUNCT bracis-19071 101 27 width width ADJ bracis-19071 101 28 , , PUNCT bracis-19071 101 29 and and CCONJ bracis-19071 101 30 resolution resolution NOUN bracis-19071 101 31 . . PUNCT bracis-19071 102 1 inceptionv3 inceptionv3 NOUN bracis-19071 102 2 . . PUNCT bracis-19071 103 1 this this DET bracis-19071 103 2 architecture architecture NOUN bracis-19071 103 3 is be AUX bracis-19071 103 4 a a DET bracis-19071 103 5 refinement refinement NOUN bracis-19071 103 6 of of ADP bracis-19071 103 7 its its PRON bracis-19071 103 8 antecessors antecessor NOUN bracis-19071 103 9 , , PUNCT bracis-19071 103 10 firstly firstly ADV bracis-19071 103 11 by by ADP bracis-19071 103 12 the the DET bracis-19071 103 13 introduction introduction NOUN bracis-19071 103 14 of of ADP bracis-19071 103 15 batch batch NOUN bracis-19071 103 16 normalization normalization NOUN bracis-19071 103 17 , , PUNCT bracis-19071 103 18 and and CCONJ bracis-19071 103 19 later later ADV bracis-19071 103 20 by by ADP bracis-19071 103 21 additional additional ADJ bracis-19071 103 22 factorization factorization NOUN bracis-19071 103 23 ideas idea NOUN bracis-19071 103 24 in in ADP bracis-19071 103 25 the the DET bracis-19071 103 26 third third ADJ bracis-19071 103 27 iteration iteration NOUN bracis-19071 103 28 [ [ X bracis-19071 103 29 14 14 NUM bracis-19071 103 30 ] ] PUNCT bracis-19071 103 31 . . PUNCT bracis-19071 104 1 mobilenetv2 mobilenetv2 PROPN bracis-19071 104 2 . . PUNCT bracis-19071 105 1 this this DET bracis-19071 105 2 network network NOUN bracis-19071 105 3 uses use VERB bracis-19071 105 4 light light ADJ bracis-19071 105 5 convolution convolution NOUN bracis-19071 105 6 layers layer NOUN bracis-19071 105 7 to to PART bracis-19071 105 8 filter filter VERB bracis-19071 105 9 the the DET bracis-19071 105 10 features feature NOUN bracis-19071 105 11 in in ADP bracis-19071 105 12 the the DET bracis-19071 105 13 intermediate intermediate ADJ bracis-19071 105 14 layer layer NOUN bracis-19071 105 15 , , PUNCT bracis-19071 105 16 is be AUX bracis-19071 105 17 based base VERB bracis-19071 105 18 on on ADP bracis-19071 105 19 a a DET bracis-19071 105 20 residual residual ADJ bracis-19071 105 21 inverted inverted ADJ bracis-19071 105 22 structure structure NOUN bracis-19071 105 23 [ [ X bracis-19071 105 24 13 13 NUM bracis-19071 105 25 ] ] PUNCT bracis-19071 105 26 . . PUNCT bracis-19071 106 1 resnet50 resnet50 PROPN bracis-19071 106 2 . . PUNCT bracis-19071 107 1 resnet resnet VERB bracis-19071 108 1 [ [ X bracis-19071 108 2 6 6 NUM bracis-19071 108 3 ] ] PUNCT bracis-19071 108 4 is be AUX bracis-19071 108 5 an an DET bracis-19071 108 6 abbreviation abbreviation NOUN bracis-19071 108 7 for for ADP bracis-19071 108 8 residual residual ADJ bracis-19071 108 9 networks network NOUN bracis-19071 108 10 . . PUNCT bracis-19071 109 1 this this DET bracis-19071 109 2 type type NOUN bracis-19071 109 3 of of ADP bracis-19071 109 4 deep deep ADJ bracis-19071 109 5 convolutional convolutional ADJ bracis-19071 109 6 neural neural ADJ bracis-19071 109 7 network network NOUN bracis-19071 109 8 uses use VERB bracis-19071 109 9 residual residual ADJ bracis-19071 109 10 blocks block NOUN bracis-19071 109 11 and and CCONJ bracis-19071 109 12 can can AUX bracis-19071 109 13 work work VERB bracis-19071 109 14 with with ADP bracis-19071 109 15 many many ADJ bracis-19071 109 16 layers layer NOUN bracis-19071 109 17 of of ADP bracis-19071 109 18 depths depth NOUN bracis-19071 109 19 , , PUNCT bracis-19071 109 20 avoiding avoid VERB bracis-19071 109 21 the the DET bracis-19071 109 22 vanishing vanish VERB bracis-19071 109 23 gradient gradient NOUN bracis-19071 109 24 problem problem NOUN bracis-19071 109 25 . . PUNCT bracis-19071 110 1 specifically specifically ADV bracis-19071 110 2 , , PUNCT bracis-19071 110 3 resnet50 resnet50 NOUN bracis-19071 110 4 is be AUX bracis-19071 110 5 a a DET bracis-19071 110 6 50 50 NUM bracis-19071 110 7 - - PUNCT bracis-19071 110 8 layer layer NOUN bracis-19071 110 9 residual residual ADJ bracis-19071 110 10 network network NOUN bracis-19071 110 11 . . PUNCT bracis-19071 111 1 xception xception PROPN bracis-19071 111 2 . . PUNCT bracis-19071 112 1 xception xception NOUN bracis-19071 113 1 [ [ X bracis-19071 113 2 2 2 X bracis-19071 113 3 ] ] PUNCT bracis-19071 113 4 is be AUX bracis-19071 113 5 a a DET bracis-19071 113 6 novel novel ADJ bracis-19071 113 7 deep deep ADJ bracis-19071 113 8 convolutional convolutional ADJ bracis-19071 113 9 neural neural ADJ bracis-19071 113 10 network network NOUN bracis-19071 113 11 architecture architecture NOUN bracis-19071 113 12 inspired inspire VERB bracis-19071 113 13 by by ADP bracis-19071 113 14 inception inception NOUN bracis-19071 113 15 [ [ X bracis-19071 113 16 14 14 NUM bracis-19071 113 17 ] ] PUNCT bracis-19071 113 18 , , PUNCT bracis-19071 113 19 where where SCONJ bracis-19071 113 20 inception inception ADJ bracis-19071 113 21 modules module NOUN bracis-19071 113 22 have have AUX bracis-19071 113 23 been be AUX bracis-19071 113 24 replaced replace VERB bracis-19071 113 25 with with ADP bracis-19071 113 26 depthwise depthwise NOUN bracis-19071 113 27 separable separable ADJ bracis-19071 113 28 convolutions convolution NOUN bracis-19071 113 29 . . PUNCT bracis-19071 114 1 the the DET bracis-19071 114 2 models model NOUN bracis-19071 114 3 were be AUX bracis-19071 114 4 evaluated evaluate VERB bracis-19071 114 5 using use VERB bracis-19071 114 6 5 5 NUM bracis-19071 114 7 - - ADJ bracis-19071 114 8 fold fold ADJ bracis-19071 114 9 stratified stratified ADJ bracis-19071 114 10 cross cross NOUN bracis-19071 114 11 - - NOUN bracis-19071 114 12 validation validation ADJ bracis-19071 114 13 . . PUNCT bracis-19071 115 1 table table NOUN bracis-19071 115 2     SPACE bracis-19071 115 3 5 5 NUM bracis-19071 115 4 shows show VERB bracis-19071 115 5 that that SCONJ bracis-19071 115 6 all all DET bracis-19071 115 7 cnns cnn NOUN bracis-19071 115 8 outperformed outperform VERB bracis-19071 115 9 almost almost ADV bracis-19071 115 10 all all PRON bracis-19071 115 11 classic classic ADJ bracis-19071 115 12 algorithms algorithm NOUN bracis-19071 115 13 , , PUNCT bracis-19071 115 14 as as SCONJ bracis-19071 115 15 expected expect VERB bracis-19071 115 16 . . PUNCT bracis-19071 116 1 most most ADJ bracis-19071 116 2 architectures architecture NOUN bracis-19071 116 3 achieved achieve VERB bracis-19071 116 4 accuracies accuracy NOUN bracis-19071 116 5 over over ADP bracis-19071 116 6 0.74 0.74 NUM bracis-19071 116 7 , , PUNCT bracis-19071 116 8 except except SCONJ bracis-19071 116 9 for for ADP bracis-19071 116 10 efficientnetb3 efficientnetb3 PROPN bracis-19071 116 11 , , PUNCT bracis-19071 116 12 which which PRON bracis-19071 116 13 achieved achieve VERB bracis-19071 116 14 the the DET bracis-19071 116 15 worst bad ADJ bracis-19071 116 16 performance performance NOUN bracis-19071 116 17 among among ADP bracis-19071 116 18 the the DET bracis-19071 116 19 deep deep ADJ bracis-19071 116 20 learning learning NOUN bracis-19071 116 21 architectures architecture NOUN bracis-19071 116 22 . . PUNCT bracis-19071 117 1 results result NOUN bracis-19071 117 2 also also ADV bracis-19071 117 3 show show VERB bracis-19071 117 4 that that SCONJ bracis-19071 117 5 mobilenetv2 mobilenetv2 PROPN bracis-19071 117 6 holds hold VERB bracis-19071 117 7 the the DET bracis-19071 117 8 highest high ADJ bracis-19071 117 9 accuracy accuracy NOUN bracis-19071 117 10 and and CCONJ bracis-19071 117 11 can can AUX bracis-19071 117 12 be be AUX bracis-19071 117 13 considered consider VERB bracis-19071 117 14 the the DET bracis-19071 117 15 state state NOUN bracis-19071 117 16 of of ADP bracis-19071 117 17 the the DET bracis-19071 117 18 art art NOUN bracis-19071 117 19 for for ADP bracis-19071 117 20 the the DET bracis-19071 117 21 iris iris NOUN bracis-19071 117 22 - - PUNCT bracis-19071 117 23 cv cv PROPN bracis-19071 117 24 dataset dataset NOUN bracis-19071 117 25 . . PUNCT bracis-19071 118 1 table table NOUN bracis-19071 118 2 5 5 NUM bracis-19071 118 3 . . PUNCT bracis-19071 118 4 results result NOUN bracis-19071 118 5 with with ADP bracis-19071 118 6 standard standard ADJ bracis-19071 118 7 deviation deviation NOUN bracis-19071 118 8 ( ( PUNCT bracis-19071 118 9 std std NOUN bracis-19071 118 10 ) ) PUNCT bracis-19071 118 11 of of ADP bracis-19071 118 12 deep deep ADJ bracis-19071 118 13 learning learning NOUN bracis-19071 118 14 architectures architecture NOUN bracis-19071 118 15 for for ADP bracis-19071 118 16 the the DET bracis-19071 118 17 iris iris NOUN bracis-19071 118 18 - - PUNCT bracis-19071 118 19 cv cv PROPN bracis-19071 118 20 dataset dataset NOUN bracis-19071 118 21 using use VERB bracis-19071 118 22 5 5 NUM bracis-19071 118 23 - - ADJ bracis-19071 118 24 fold fold ADJ bracis-19071 118 25 cross cross NOUN bracis-19071 118 26 - - NOUN bracis-19071 118 27 validation validation NOUN bracis-19071 118 28 . . PUNCT bracis-19071 119 1 columns column NOUN bracis-19071 119 2 b b PROPN bracis-19071 119 3 and and CCONJ bracis-19071 119 4 e e NOUN bracis-19071 119 5 refer refer VERB bracis-19071 119 6 to to PART bracis-19071 119 7 batch batch VERB bracis-19071 119 8 size size NOUN bracis-19071 119 9 and and CCONJ bracis-19071 119 10 number number NOUN bracis-19071 119 11 of of ADP bracis-19071 119 12 epochs epoch NOUN bracis-19071 119 13 , , PUNCT bracis-19071 119 14 respectively respectively ADV bracis-19071 119 15 . . PUNCT bracis-19071 120 1 we we PRON bracis-19071 120 2 used use VERB bracis-19071 120 3 different different ADJ bracis-19071 120 4 batch batch NOUN bracis-19071 120 5 sizes size NOUN bracis-19071 120 6 considering consider VERB bracis-19071 120 7 hardware hardware NOUN bracis-19071 120 8 resources resource NOUN bracis-19071 120 9 of of ADP bracis-19071 120 10 each each DET bracis-19071 120 11 machine machine NOUN bracis-19071 120 12 . . PUNCT bracis-19071 121 1 the the DET bracis-19071 121 2 densenet121 densenet121 PROPN bracis-19071 121 3 , , PUNCT bracis-19071 121 4 efficientnetb0 efficientnetb0 PROPN bracis-19071 121 5 , , PUNCT bracis-19071 121 6 and and CCONJ bracis-19071 121 7 mobilenetv2 mobilenetv2 PROPN bracis-19071 121 8 architectures architecture NOUN bracis-19071 121 9 were be AUX bracis-19071 121 10 trained train VERB bracis-19071 121 11 using use VERB bracis-19071 121 12 rtx rtx PROPN bracis-19071 121 13 2060 2060 NUM bracis-19071 121 14 super super PROPN bracis-19071 121 15 , , PUNCT bracis-19071 121 16 resnet resnet VERB bracis-19071 121 17 50 50 NUM bracis-19071 121 18 was be AUX bracis-19071 121 19 trained train VERB bracis-19071 121 20 with with ADP bracis-19071 121 21 rtx rtx PROPN bracis-19071 121 22 2070 2070 NUM bracis-19071 121 23 , , PUNCT bracis-19071 121 24 while while SCONJ bracis-19071 121 25 the the DET bracis-19071 121 26 remaining remain VERB bracis-19071 121 27 ones one NOUN bracis-19071 121 28 were be AUX bracis-19071 121 29 trained train VERB bracis-19071 121 30 with with ADP bracis-19071 121 31 a a DET bracis-19071 121 32 gtx gtx ADJ bracis-19071 121 33 1660ti.full 1660ti.full NUM bracis-19071 121 34 size size NOUN bracis-19071 121 35 table table NOUN bracis-19071 121 36 in in ADP bracis-19071 121 37 addition addition NOUN bracis-19071 121 38 to to ADP bracis-19071 121 39 assessing assess VERB bracis-19071 121 40 accuracy accuracy NOUN bracis-19071 121 41 , , PUNCT bracis-19071 121 42 we we PRON bracis-19071 121 43 analyzed analyze VERB bracis-19071 121 44 the the DET bracis-19071 121 45 confusion confusion NOUN bracis-19071 121 46 matrix matrix NOUN bracis-19071 121 47 that that PRON bracis-19071 121 48 corresponds correspond VERB bracis-19071 121 49 to to ADP bracis-19071 121 50 the the DET bracis-19071 121 51 best good ADJ bracis-19071 121 52 mobilenetv2 mobilenetv2 NOUN bracis-19071 121 53 result result VERB bracis-19071 121 54 , , PUNCT bracis-19071 121 55 so so SCONJ bracis-19071 121 56 that that SCONJ bracis-19071 121 57 we we PRON bracis-19071 121 58 can can AUX bracis-19071 121 59 determine determine VERB bracis-19071 121 60 which which DET bracis-19071 121 61 classes class VERB bracis-19071 121 62 the the DET bracis-19071 121 63 models model NOUN bracis-19071 121 64 have have VERB bracis-19071 121 65 the the DET bracis-19071 121 66 most most ADJ bracis-19071 121 67 difficulty difficulty NOUN bracis-19071 121 68 in in ADP bracis-19071 121 69 predicting predict VERB bracis-19071 121 70 . . PUNCT bracis-19071 122 1 the the DET bracis-19071 122 2 confusion confusion NOUN bracis-19071 122 3 matrix matrix NOUN bracis-19071 122 4 is be AUX bracis-19071 122 5 shown show VERB bracis-19071 122 6 in in ADP bracis-19071 122 7 table table NOUN bracis-19071 122 8     SPACE bracis-19071 122 9 6 6 NUM bracis-19071 122 10 . . PUNCT bracis-19071 122 11 table table NOUN bracis-19071 122 12 6 6 NUM bracis-19071 122 13 . . PUNCT bracis-19071 122 14 confusion confusion NOUN bracis-19071 122 15 matrix matrix NOUN bracis-19071 122 16 corresponding correspond VERB bracis-19071 122 17 to to ADP bracis-19071 122 18 mobilenetv2 mobilenetv2 PROPN bracis-19071 122 19 ’s ’s PART bracis-19071 122 20 best good ADJ bracis-19071 122 21 test test NOUN bracis-19071 122 22 accuracy accuracy NOUN bracis-19071 122 23 – – PUNCT bracis-19071 122 24 each each DET bracis-19071 122 25 class class NOUN bracis-19071 122 26 is be AUX bracis-19071 122 27 represented represent VERB bracis-19071 122 28 by by ADP bracis-19071 122 29 the the DET bracis-19071 122 30 first first ADJ bracis-19071 122 31 two two NUM bracis-19071 122 32 letters letter NOUN bracis-19071 122 33 in in ADP bracis-19071 122 34 its its PRON bracis-19071 122 35 common common ADJ bracis-19071 122 36 name.full name.full ADJ bracis-19071 122 37 size size NOUN bracis-19071 122 38 table table NOUN bracis-19071 122 39 the the DET bracis-19071 122 40 mobilenetv2 mobilenetv2 PROPN bracis-19071 122 41 confusion confusion NOUN bracis-19071 122 42 matrix matrix NOUN bracis-19071 122 43 shows show VERB bracis-19071 122 44 that that SCONJ bracis-19071 122 45 the the DET bracis-19071 122 46 network network NOUN bracis-19071 122 47 was be AUX bracis-19071 122 48 good good ADJ bracis-19071 122 49 at at ADP bracis-19071 122 50 differentiating differentiate VERB bracis-19071 122 51 western western ADJ bracis-19071 122 52 iris iris NOUN bracis-19071 122 53 ( ( PUNCT bracis-19071 122 54 we we NOUN bracis-19071 122 55 ) ) PUNCT bracis-19071 122 56 , , PUNCT bracis-19071 122 57 yellow yellow ADJ bracis-19071 122 58 iris iris NOUN bracis-19071 122 59 ( ( PUNCT bracis-19071 122 60 ye ye NOUN bracis-19071 122 61 ) ) PUNCT bracis-19071 122 62 , , PUNCT bracis-19071 122 63 and and CCONJ bracis-19071 122 64 dwarf dwarf NOUN bracis-19071 122 65 crested crest VERB bracis-19071 122 66 iris iris NOUN bracis-19071 122 67 ( ( PUNCT bracis-19071 122 68 dw dw NOUN bracis-19071 122 69 ) ) PUNCT bracis-19071 122 70 from from ADP bracis-19071 122 71 the the DET bracis-19071 122 72 other other ADJ bracis-19071 122 73 classes class NOUN bracis-19071 122 74 , , PUNCT bracis-19071 122 75 as as SCONJ bracis-19071 122 76 their their PRON bracis-19071 122 77 precision precision NOUN bracis-19071 122 78 scores score NOUN bracis-19071 122 79 corresponded correspond VERB bracis-19071 122 80 to to ADP bracis-19071 122 81 approximately approximately ADV bracis-19071 122 82 88 88 NUM bracis-19071 122 83 % % NOUN bracis-19071 122 84 , , PUNCT bracis-19071 122 85 89 89 NUM bracis-19071 122 86 % % NOUN bracis-19071 122 87 , , PUNCT bracis-19071 122 88 and and CCONJ bracis-19071 122 89 89 89 NUM bracis-19071 122 90 % % NOUN bracis-19071 122 91 , , PUNCT bracis-19071 122 92 respectively respectively ADV bracis-19071 122 93 , , PUNCT bracis-19071 122 94 as as SCONJ bracis-19071 122 95 shown show VERB bracis-19071 122 96 in in ADP bracis-19071 122 97 table table NOUN bracis-19071 122 98     SPACE bracis-19071 122 99 7 7 NUM bracis-19071 122 100 . . PUNCT bracis-19071 123 1 the the DET bracis-19071 123 2 other other ADJ bracis-19071 123 3 two two NUM bracis-19071 123 4 classes class NOUN bracis-19071 123 5 – – PUNCT bracis-19071 123 6 bearded bearded ADJ bracis-19071 123 7 iris iris NOUN bracis-19071 123 8 ( ( PUNCT bracis-19071 123 9 be be AUX bracis-19071 123 10 ) ) PUNCT bracis-19071 123 11 and and CCONJ bracis-19071 123 12 douglas douglas PROPN bracis-19071 123 13 iris iris PROPN bracis-19071 123 14 ( ( PUNCT bracis-19071 123 15 do do PROPN bracis-19071 123 16 ) ) PUNCT bracis-19071 123 17 – – PUNCT bracis-19071 123 18 were be AUX bracis-19071 123 19 not not PART bracis-19071 123 20 as as ADV bracis-19071 123 21 well well ADV bracis-19071 123 22 discriminated discriminate VERB bracis-19071 123 23 , , PUNCT bracis-19071 123 24 and and CCONJ bracis-19071 123 25 the the DET bracis-19071 123 26 algorithm algorithm NOUN bracis-19071 123 27 had have VERB bracis-19071 123 28 some some DET bracis-19071 123 29 trouble trouble NOUN bracis-19071 123 30 distinguishing distinguish VERB bracis-19071 123 31 them they PRON bracis-19071 123 32 correctly correctly ADV bracis-19071 123 33 since since SCONJ bracis-19071 123 34 the the DET bracis-19071 123 35 flowers flower NOUN bracis-19071 123 36 belonging belong VERB bracis-19071 123 37 to to ADP bracis-19071 123 38 these these DET bracis-19071 123 39 two two NUM bracis-19071 123 40 classes class NOUN bracis-19071 123 41 often often ADV bracis-19071 123 42 have have VERB bracis-19071 123 43 the the DET bracis-19071 123 44 same same ADJ bracis-19071 123 45 color color NOUN bracis-19071 123 46 , , PUNCT bracis-19071 123 47 although although SCONJ bracis-19071 123 48 they they PRON bracis-19071 123 49 have have VERB bracis-19071 123 50 different different ADJ bracis-19071 123 51 petal petal ADJ bracis-19071 123 52 shapes shape NOUN bracis-19071 123 53 . . PUNCT bracis-19071 124 1 the the DET bracis-19071 124 2 similarity similarity NOUN bracis-19071 124 3 between between ADP bracis-19071 124 4 these these DET bracis-19071 124 5 two two NUM bracis-19071 124 6 classes class NOUN bracis-19071 124 7 can can AUX bracis-19071 124 8 be be AUX bracis-19071 124 9 observed observe VERB bracis-19071 124 10 in in ADP bracis-19071 124 11 table table NOUN bracis-19071 124 12     SPACE bracis-19071 124 13 8 8 NUM bracis-19071 124 14 . . PUNCT bracis-19071 125 1 douglas douglas PROPN bracis-19071 125 2 iris iris PROPN bracis-19071 125 3 can can AUX bracis-19071 125 4 also also ADV bracis-19071 125 5 sometimes sometimes ADV bracis-19071 125 6 be be AUX bracis-19071 125 7 confused confuse VERB bracis-19071 125 8 with with ADP bracis-19071 125 9 dwarf dwarf NOUN bracis-19071 125 10 crested crest VERB bracis-19071 125 11 irises iris NOUN bracis-19071 125 12 , , PUNCT bracis-19071 125 13 due due ADP bracis-19071 125 14 to to ADP bracis-19071 125 15 their their PRON bracis-19071 125 16 white white ADJ bracis-19071 125 17 petal petal ADJ bracis-19071 125 18 markings marking NOUN bracis-19071 125 19 . . PUNCT bracis-19071 126 1 as as ADP bracis-19071 126 2 a a DET bracis-19071 126 3 result result NOUN bracis-19071 126 4 , , PUNCT bracis-19071 126 5 douglas douglas PROPN bracis-19071 126 6 irises iris NOUN bracis-19071 126 7 were be AUX bracis-19071 126 8 the the DET bracis-19071 126 9 hardest hard ADJ bracis-19071 126 10 flowers flower NOUN bracis-19071 126 11 to to PART bracis-19071 126 12 classify classify VERB bracis-19071 126 13 correctly correctly ADV bracis-19071 126 14 , , PUNCT bracis-19071 126 15 with with ADP bracis-19071 126 16 only only ADV bracis-19071 126 17 69 69 NUM bracis-19071 126 18 % % NOUN bracis-19071 126 19 precision precision NOUN bracis-19071 126 20 . . PUNCT bracis-19071 127 1 table table NOUN bracis-19071 127 2 7 7 NUM bracis-19071 127 3 . . PUNCT bracis-19071 127 4 precision precision NOUN bracis-19071 127 5 , , PUNCT bracis-19071 127 6 recall recall NOUN bracis-19071 127 7 and and CCONJ bracis-19071 127 8 f1 f1 NOUN bracis-19071 127 9 - - PUNCT bracis-19071 127 10 score score NOUN bracis-19071 127 11 corresponding corresponding NOUN bracis-19071 127 12 to to ADP bracis-19071 127 13 mobilenetv2 mobilenetv2 PROPN bracis-19071 127 14 ’s ’s PART bracis-19071 127 15 best good ADJ bracis-19071 127 16 results result NOUN bracis-19071 127 17 in in ADP bracis-19071 127 18 test test NOUN bracis-19071 127 19 dataset dataset VERB bracis-19071 127 20 each each DET bracis-19071 127 21 class class NOUN bracis-19071 127 22 is be AUX bracis-19071 127 23 represented represent VERB bracis-19071 127 24 by by ADP bracis-19071 127 25 the the DET bracis-19071 127 26 first first ADJ bracis-19071 127 27 letters letter NOUN bracis-19071 127 28 in in ADP bracis-19071 127 29 its its PRON bracis-19071 127 30 common common ADJ bracis-19071 127 31 namefull namefull ADJ bracis-19071 127 32 size size NOUN bracis-19071 127 33 table table NOUN bracis-19071 127 34 table table NOUN bracis-19071 127 35 8 8 NUM bracis-19071 127 36 . . PUNCT bracis-19071 128 1 examples example NOUN bracis-19071 128 2 of of ADP bracis-19071 128 3 bearded bearded ADJ bracis-19071 128 4 iris iris NOUN bracis-19071 128 5 and and CCONJ bracis-19071 128 6 douglas douglas PROPN bracis-19071 128 7 iris iris NOUN bracis-19071 128 8 flowers flower NOUN bracis-19071 128 9 showing show VERB bracis-19071 128 10 their their PRON bracis-19071 128 11 similar similar ADJ bracis-19071 128 12 colors color NOUN bracis-19071 128 13 , , PUNCT bracis-19071 128 14 but but CCONJ bracis-19071 128 15 different different ADJ bracis-19071 128 16 petal petal ADJ bracis-19071 128 17 structures.full structures.full NOUN bracis-19071 128 18 size size NOUN bracis-19071 128 19 table table NOUN bracis-19071 128 20 5 5 NUM bracis-19071 128 21 conclusion conclusion NOUN bracis-19071 128 22 this this DET bracis-19071 128 23 paper paper NOUN bracis-19071 128 24 introduced introduce VERB bracis-19071 128 25 a a DET bracis-19071 128 26 new new ADJ bracis-19071 128 27 computer computer NOUN bracis-19071 128 28 vision vision NOUN bracis-19071 128 29 dataset dataset NOUN bracis-19071 128 30 , , PUNCT bracis-19071 128 31 called call VERB bracis-19071 128 32 iris iris NOUN bracis-19071 128 33 - - PUNCT bracis-19071 128 34 cv cv PROPN bracis-19071 128 35 , , PUNCT bracis-19071 128 36 consisting consist VERB bracis-19071 128 37 of of ADP bracis-19071 128 38 five five NUM bracis-19071 128 39 classes class NOUN bracis-19071 128 40 of of ADP bracis-19071 128 41 iris iris NOUN bracis-19071 128 42 flowers flower NOUN bracis-19071 128 43 . . PUNCT bracis-19071 129 1 the the DET bracis-19071 129 2 images image NOUN bracis-19071 129 3 show show VERB bracis-19071 129 4 many many ADJ bracis-19071 129 5 features feature NOUN bracis-19071 129 6 that that PRON bracis-19071 129 7 make make VERB bracis-19071 129 8 this this PRON bracis-19071 129 9 dataset dataset NOUN bracis-19071 129 10 a a DET bracis-19071 129 11 challenging challenging ADJ bracis-19071 129 12 task task NOUN bracis-19071 129 13 , , PUNCT bracis-19071 129 14 such such ADJ bracis-19071 129 15 as as ADP bracis-19071 129 16 non non ADJ bracis-19071 129 17 - - ADJ bracis-19071 129 18 centered centered ADJ bracis-19071 129 19 flowers flower NOUN bracis-19071 129 20 , , PUNCT bracis-19071 129 21 different different ADJ bracis-19071 129 22 lighting lighting NOUN bracis-19071 129 23 conditions condition NOUN bracis-19071 129 24 , , PUNCT bracis-19071 129 25 multiple multiple ADJ bracis-19071 129 26 flowers flower NOUN bracis-19071 129 27 per per ADP bracis-19071 129 28 image image NOUN bracis-19071 129 29 , , PUNCT bracis-19071 129 30 and and CCONJ bracis-19071 129 31 classes class NOUN bracis-19071 129 32 that that PRON bracis-19071 129 33 naturally naturally ADV bracis-19071 129 34 appear appear VERB bracis-19071 129 35 with with ADP bracis-19071 129 36 different different ADJ bracis-19071 129 37 petal petal ADJ bracis-19071 129 38 colors color NOUN bracis-19071 129 39 . . PUNCT bracis-19071 130 1 due due ADP bracis-19071 130 2 to to ADP bracis-19071 130 3 all all PRON bracis-19071 130 4 of of ADP bracis-19071 130 5 these these DET bracis-19071 130 6 reasons reason NOUN bracis-19071 130 7 , , PUNCT bracis-19071 130 8 iris iris PROPN bracis-19071 130 9 - - PUNCT bracis-19071 130 10 cv cv PROPN bracis-19071 130 11 proved prove VERB bracis-19071 130 12 to to PART bracis-19071 130 13 be be AUX bracis-19071 130 14 too too ADV bracis-19071 130 15 hard hard ADJ bracis-19071 130 16 for for ADP bracis-19071 130 17 traditional traditional ADJ bracis-19071 130 18 machine machine NOUN bracis-19071 130 19 learning learning NOUN bracis-19071 130 20 algorithms algorithm NOUN bracis-19071 130 21 , , PUNCT bracis-19071 130 22 with with ADP bracis-19071 130 23 poorer poor ADJ bracis-19071 130 24 results result NOUN bracis-19071 130 25 than than ADP bracis-19071 130 26 those those PRON bracis-19071 130 27 observed observe VERB bracis-19071 130 28 for for ADP bracis-19071 130 29 established establish VERB bracis-19071 130 30 benchmark benchmark NOUN bracis-19071 130 31 datasets dataset NOUN bracis-19071 130 32 , , PUNCT bracis-19071 130 33 such such ADJ bracis-19071 130 34 as as ADP bracis-19071 130 35 mnist mnist NOUN bracis-19071 130 36 and and CCONJ bracis-19071 130 37 fashion fashion NOUN bracis-19071 130 38 - - PUNCT bracis-19071 130 39 mnist mnist NOUN bracis-19071 130 40 . . PUNCT bracis-19071 131 1 state state NOUN bracis-19071 131 2 - - PUNCT bracis-19071 131 3 of of ADP bracis-19071 131 4 - - PUNCT bracis-19071 131 5 the the DET bracis-19071 131 6 - - PUNCT bracis-19071 131 7 art art NOUN bracis-19071 131 8 deep deep ADJ bracis-19071 131 9 neural neural ADJ bracis-19071 131 10 nets net NOUN bracis-19071 131 11 also also ADV bracis-19071 131 12 performed perform VERB bracis-19071 131 13 worse bad ADJ bracis-19071 131 14 than than ADP bracis-19071 131 15 current current ADJ bracis-19071 131 16 results result NOUN bracis-19071 131 17 for for ADP bracis-19071 131 18 mnist mnist NOUN bracis-19071 131 19 , , PUNCT bracis-19071 131 20 fashion fashion NOUN bracis-19071 131 21 - - PUNCT bracis-19071 131 22 mnist mnist NOUN bracis-19071 131 23 , , PUNCT bracis-19071 131 24 and and CCONJ bracis-19071 131 25 imagenet imagenet NOUN bracis-19071 131 26 , , PUNCT bracis-19071 131 27 with with ADP bracis-19071 131 28 mobilenetv2 mobilenetv2 PROPN bracis-19071 131 29 achieving achieve VERB bracis-19071 131 30 82 82 NUM bracis-19071 131 31 % % NOUN bracis-19071 131 32 accuracy accuracy NOUN bracis-19071 131 33 , , PUNCT bracis-19071 131 34 which which PRON bracis-19071 131 35 is be AUX bracis-19071 131 36 the the DET bracis-19071 131 37 best good ADJ bracis-19071 131 38 cross cross ADJ bracis-19071 131 39 - - ADJ bracis-19071 131 40 validated validated ADJ bracis-19071 131 41 result result NOUN bracis-19071 131 42 so so ADV bracis-19071 131 43 far far ADV bracis-19071 131 44 . . PUNCT bracis-19071 132 1 additionally additionally ADV bracis-19071 132 2 , , PUNCT bracis-19071 132 3 an an DET bracis-19071 132 4 analysis analysis NOUN bracis-19071 132 5 of of ADP bracis-19071 132 6 the the DET bracis-19071 132 7 best good ADJ bracis-19071 132 8 confusion confusion NOUN bracis-19071 132 9 matrix matrix NOUN bracis-19071 132 10 produced produce VERB bracis-19071 132 11 by by ADP bracis-19071 132 12 mobilenetv2 mobilenetv2 PROPN bracis-19071 132 13 showed show VERB bracis-19071 132 14 that that SCONJ bracis-19071 132 15 three three NUM bracis-19071 132 16 of of ADP bracis-19071 132 17 the the DET bracis-19071 132 18 classes class NOUN bracis-19071 132 19 are be AUX bracis-19071 132 20 more more ADV bracis-19071 132 21 easily easily ADV bracis-19071 132 22 classified classified ADJ bracis-19071 132 23 , , PUNCT bracis-19071 132 24 namely namely ADV bracis-19071 132 25 dwarf dwarf VERB bracis-19071 132 26 crested crested ADJ bracis-19071 132 27 iris iris NOUN bracis-19071 132 28 , , PUNCT bracis-19071 132 29 western western ADJ bracis-19071 132 30 iris iris NOUN bracis-19071 132 31 , , PUNCT bracis-19071 132 32 and and CCONJ bracis-19071 132 33 yellow yellow ADJ bracis-19071 132 34 iris iris NOUN bracis-19071 132 35 . . PUNCT bracis-19071 133 1 the the DET bracis-19071 133 2 remaining remain VERB bracis-19071 133 3 two two NUM bracis-19071 133 4 classes class NOUN bracis-19071 133 5 – – PUNCT bracis-19071 133 6 bearded bearded ADJ bracis-19071 133 7 iris iris NOUN bracis-19071 133 8 and and CCONJ bracis-19071 133 9 douglas douglas PROPN bracis-19071 133 10 iris iris PROPN bracis-19071 133 11 – – PUNCT bracis-19071 133 12 offered offer VERB bracis-19071 133 13 harder hard ADJ bracis-19071 133 14 challenges challenge NOUN bracis-19071 133 15 , , PUNCT bracis-19071 133 16 with with ADP bracis-19071 133 17 the the DET bracis-19071 133 18 latter latter ADJ bracis-19071 133 19 being be AUX bracis-19071 133 20 the the DET bracis-19071 133 21 toughest tough ADJ bracis-19071 133 22 to to PART bracis-19071 133 23 discriminate discriminate VERB bracis-19071 133 24 . . PUNCT bracis-19071 134 1 since since SCONJ bracis-19071 134 2 this this DET bracis-19071 134 3 paper paper NOUN bracis-19071 134 4 main main ADJ bracis-19071 134 5 objective objective NOUN bracis-19071 134 6 is be AUX bracis-19071 134 7 to to PART bracis-19071 134 8 propose propose VERB bracis-19071 134 9 a a DET bracis-19071 134 10 new new ADJ bracis-19071 134 11 iris iris NOUN bracis-19071 134 12 dataset dataset VERB bracis-19071 134 13 considering consider VERB bracis-19071 134 14 computer computer NOUN bracis-19071 134 15 vision vision NOUN bracis-19071 134 16 common common ADJ bracis-19071 134 17 problems problem NOUN bracis-19071 134 18 ( ( PUNCT bracis-19071 134 19 e.g. e.g. ADV bracis-19071 134 20 occlusion occlusion NOUN bracis-19071 134 21 , , PUNCT bracis-19071 134 22 background background NOUN bracis-19071 134 23 noise noise NOUN bracis-19071 134 24 and and CCONJ bracis-19071 134 25 fine fine ADJ bracis-19071 134 26 grained grain VERB bracis-19071 134 27 features feature NOUN bracis-19071 134 28 ) ) PUNCT bracis-19071 134 29 , , PUNCT bracis-19071 134 30 we we PRON bracis-19071 134 31 performed perform VERB bracis-19071 134 32 a a DET bracis-19071 134 33 baseline baseline NOUN bracis-19071 134 34 of of ADP bracis-19071 134 35 most most ADV bracis-19071 134 36 used use VERB bracis-19071 134 37 and and CCONJ bracis-19071 134 38 state state NOUN bracis-19071 134 39 - - PUNCT bracis-19071 134 40 of of ADP bracis-19071 134 41 - - PUNCT bracis-19071 134 42 the the DET bracis-19071 134 43 - - PUNCT bracis-19071 134 44 art art NOUN bracis-19071 134 45 algorithms algorithm NOUN bracis-19071 134 46 . . PUNCT bracis-19071 135 1 future future ADJ bracis-19071 135 2 works work NOUN bracis-19071 135 3 include include VERB bracis-19071 135 4 further further ADJ bracis-19071 135 5 exploration exploration NOUN bracis-19071 135 6 of of ADP bracis-19071 135 7 deep deep ADJ bracis-19071 135 8 neural neural ADJ bracis-19071 135 9 net net ADJ bracis-19071 135 10 architectures architecture NOUN bracis-19071 135 11 and and CCONJ bracis-19071 135 12 regularization regularization NOUN bracis-19071 135 13 techniques technique NOUN bracis-19071 135 14 since since SCONJ bracis-19071 135 15 the the DET bracis-19071 135 16 architectures architecture NOUN bracis-19071 135 17 used use VERB bracis-19071 135 18 in in ADP bracis-19071 135 19 this this DET bracis-19071 135 20 work work NOUN bracis-19071 135 21 focused focus VERB bracis-19071 135 22 on on ADP bracis-19071 135 23 learning learn VERB bracis-19071 135 24 large large ADJ bracis-19071 135 25 datasets dataset NOUN bracis-19071 135 26 and and CCONJ bracis-19071 135 27 some some PRON bracis-19071 135 28 of of ADP bracis-19071 135 29 them they PRON bracis-19071 135 30 overfitted overfitte VERB bracis-19071 135 31 . . PUNCT bracis-19071 136 1 in in ADP bracis-19071 136 2 addition addition NOUN bracis-19071 136 3 , , PUNCT bracis-19071 136 4 hyper hyper ADJ bracis-19071 136 5 parameter parameter NOUN bracis-19071 136 6 tuning tuning NOUN bracis-19071 136 7 is be AUX bracis-19071 136 8 also also ADV bracis-19071 136 9 possible possible ADJ bracis-19071 136 10 through through ADP bracis-19071 136 11 optimization optimization NOUN bracis-19071 136 12 techniques technique NOUN bracis-19071 136 13 such such ADJ bracis-19071 136 14 as as ADP bracis-19071 136 15 bayesian bayesian NOUN bracis-19071 136 16 optimization optimization NOUN bracis-19071 136 17 , , PUNCT bracis-19071 136 18 grid grid NOUN bracis-19071 136 19 search search NOUN bracis-19071 136 20 or or CCONJ bracis-19071 136 21 random random ADJ bracis-19071 136 22 search search NOUN bracis-19071 136 23 , , PUNCT bracis-19071 136 24 specially specially ADV bracis-19071 136 25 classic classic ADJ bracis-19071 136 26 machine machine NOUN bracis-19071 136 27 learning learn VERB bracis-19071 136 28 approaches approach NOUN bracis-19071 136 29 which which PRON bracis-19071 136 30 returned return VERB bracis-19071 136 31 worse worse ADV bracis-19071 136 32 performed perform VERB bracis-19071 136 33 in in ADP bracis-19071 136 34 this this DET bracis-19071 136 35 dataset dataset NOUN bracis-19071 136 36 . . PUNCT bracis-19071 137 1 therefore therefore ADV bracis-19071 137 2 , , PUNCT bracis-19071 137 3 our our PRON bracis-19071 137 4 next next ADJ bracis-19071 137 5 goal goal NOUN bracis-19071 137 6 is be AUX bracis-19071 137 7 to to PART bracis-19071 137 8 improve improve VERB bracis-19071 137 9 the the DET bracis-19071 137 10 current current ADJ bracis-19071 137 11 benchmark benchmark NOUN bracis-19071 137 12 results result NOUN bracis-19071 137 13 , , PUNCT bracis-19071 137 14 collect collect VERB bracis-19071 137 15 more more ADJ bracis-19071 137 16 data datum NOUN bracis-19071 137 17 from from ADP bracis-19071 137 18 different different ADJ bracis-19071 137 19 sources source NOUN bracis-19071 137 20 , , PUNCT bracis-19071 137 21 and and CCONJ bracis-19071 137 22 extract extract VERB bracis-19071 137 23 an an DET bracis-19071 137 24 object object NOUN bracis-19071 137 25 detection detection NOUN bracis-19071 137 26 dataset dataset VERB bracis-19071 137 27 based base VERB bracis-19071 137 28 on on ADP bracis-19071 137 29 the the DET bracis-19071 137 30 same same ADJ bracis-19071 137 31 images image NOUN bracis-19071 137 32 . . PUNCT bracis-19071 138 1 notes note VERB bracis-19071 138 2 1.see 1.see PROPN bracis-19071 138 3 tensorflow tensorflow NOUN bracis-19071 138 4 documentation documentation NOUN bracis-19071 138 5 for for ADP bracis-19071 138 6 more more ADJ bracis-19071 138 7 details detail NOUN bracis-19071 138 8 . . PUNCT bracis-19071 139 1 references reference NOUN bracis-19071 139 2 anderson anderson PROPN bracis-19071 139 3 , , PUNCT bracis-19071 139 4 e. e. PROPN bracis-19071 139 5 : : PUNCT bracis-19071 139 6 the the DET bracis-19071 139 7 species species NOUN bracis-19071 139 8 problem problem NOUN bracis-19071 139 9 in in ADP bracis-19071 139 10 iris iris NOUN bracis-19071 139 11 . . PUNCT bracis-19071 140 1 ann ann PROPN bracis-19071 140 2 . . PROPN bracis-19071 140 3 mo mo PROPN bracis-19071 140 4 . . PROPN bracis-19071 140 5 bot bot PROPN bracis-19071 140 6 . . PUNCT bracis-19071 141 1 gard gard NOUN bracis-19071 141 2 . . PUNCT bracis-19071 142 1 23(3 23(3 NUM bracis-19071 142 2 ) ) PUNCT bracis-19071 142 3 , , PUNCT bracis-19071 142 4 457–509 457–509 NUM bracis-19071 142 5 ( ( PUNCT bracis-19071 142 6 1936 1936 NUM bracis-19071 142 7 ) ) PUNCT bracis-19071 142 8 . . PUNCT bracis-19071 143 1 http://www.jstor.org/stable/2394164 http://www.jstor.org/stable/2394164 NOUN bracis-19071 143 2 chollet chollet NOUN bracis-19071 143 3 , , PUNCT bracis-19071 143 4 f. f. PROPN bracis-19071 143 5 : : PUNCT bracis-19071 143 6 xception xception NOUN bracis-19071 143 7 : : PUNCT bracis-19071 143 8 deep deep ADJ bracis-19071 143 9 learning learn VERB bracis-19071 143 10 with with ADP bracis-19071 143 11 depthwise depthwise NOUN bracis-19071 143 12 separable separable ADJ bracis-19071 143 13 convolutions convolution NOUN bracis-19071 143 14 . . PUNCT bracis-19071 144 1 in in ADP bracis-19071 144 2 : : PUNCT bracis-19071 144 3 proceedings proceeding NOUN bracis-19071 144 4 of of ADP bracis-19071 144 5 the the DET bracis-19071 144 6 ieee ieee NOUN bracis-19071 144 7 conference conference NOUN bracis-19071 144 8 on on ADP bracis-19071 144 9 computer computer NOUN bracis-19071 144 10 vision vision NOUN bracis-19071 144 11 and and CCONJ bracis-19071 144 12 pattern pattern NOUN bracis-19071 144 13 recognition recognition NOUN bracis-19071 144 14 , , PUNCT bracis-19071 144 15 pp pp PROPN bracis-19071 144 16 . . PUNCT bracis-19071 145 1 1251–1258 1251–1258 NUM bracis-19071 145 2 ( ( PUNCT bracis-19071 145 3 2017 2017 NUM bracis-19071 145 4 ) ) PUNCT bracis-19071 145 5 google google PROPN bracis-19071 145 6 scholar scholar NOUN bracis-19071 145 7     SPACE bracis-19071 145 8 deng deng PROPN bracis-19071 145 9 , , PUNCT bracis-19071 145 10 j. j. PROPN bracis-19071 145 11 , , PUNCT bracis-19071 145 12 dong dong PROPN bracis-19071 145 13 , , PUNCT bracis-19071 145 14 w. w. PROPN bracis-19071 145 15 , , PUNCT bracis-19071 145 16 socher socher PROPN bracis-19071 145 17 , , PUNCT bracis-19071 145 18 r. r. PROPN bracis-19071 145 19 , , PUNCT bracis-19071 145 20 li li PROPN bracis-19071 145 21 , , PUNCT bracis-19071 145 22 l.j l.j PROPN bracis-19071 145 23 . . PROPN bracis-19071 145 24 , , PUNCT bracis-19071 145 25 li li PROPN bracis-19071 145 26 , , PUNCT bracis-19071 145 27 k. k. PROPN bracis-19071 145 28 , , PUNCT bracis-19071 145 29 fei fei PROPN bracis-19071 145 30 - - PUNCT bracis-19071 145 31 fei fei PROPN bracis-19071 145 32 , , PUNCT bracis-19071 145 33 l. l. PROPN bracis-19071 145 34 : : PUNCT bracis-19071 145 35 imagenet imagenet PROPN bracis-19071 145 36 : : PUNCT bracis-19071 145 37 a a DET bracis-19071 145 38 large large ADJ bracis-19071 145 39 - - PUNCT bracis-19071 145 40 scale scale NOUN bracis-19071 145 41 hierarchical hierarchical ADJ bracis-19071 145 42 image image NOUN bracis-19071 145 43 database database NOUN bracis-19071 145 44 . . PUNCT bracis-19071 146 1 in in ADP bracis-19071 146 2 : : PUNCT bracis-19071 146 3 2009 2009 NUM bracis-19071 146 4 ieee ieee NOUN bracis-19071 146 5 conference conference NOUN bracis-19071 146 6 on on ADP bracis-19071 146 7 computer computer NOUN bracis-19071 146 8 vision vision NOUN bracis-19071 146 9 and and CCONJ bracis-19071 146 10 pattern pattern NOUN bracis-19071 146 11 recognition recognition NOUN bracis-19071 146 12 , , PUNCT bracis-19071 146 13 pp pp ADP bracis-19071 146 14 . . PUNCT bracis-19071 147 1 248–255 248–255 NUM bracis-19071 147 2 . . X bracis-19071 147 3 ieee ieee NOUN bracis-19071 147 4 ( ( PUNCT bracis-19071 147 5 2009 2009 NUM bracis-19071 147 6 ) ) PUNCT bracis-19071 147 7 google google PROPN bracis-19071 147 8 scholar scholar NOUN bracis-19071 147 9     SPACE bracis-19071 147 10 fisher fisher PROPN bracis-19071 147 11 , , PUNCT bracis-19071 147 12 r.a r.a PROPN bracis-19071 147 13 . . PROPN bracis-19071 147 14 : : PUNCT bracis-19071 148 1 the the DET bracis-19071 148 2 use use NOUN bracis-19071 148 3 of of ADP bracis-19071 148 4 multiple multiple ADJ bracis-19071 148 5 measurements measurement NOUN bracis-19071 148 6 in in ADP bracis-19071 148 7 taxonomic taxonomic ADJ bracis-19071 148 8 problems problem NOUN bracis-19071 148 9 . . PUNCT bracis-19071 149 1 ann ann PROPN bracis-19071 149 2 . . PUNCT bracis-19071 150 1 eugenics eugenics PROPN bracis-19071 150 2 7(2 7(2 NUM bracis-19071 150 3 ) ) PUNCT bracis-19071 150 4 , , PUNCT bracis-19071 150 5 179–188 179–188 NUM bracis-19071 150 6 ( ( PUNCT bracis-19071 150 7 1936 1936 NUM bracis-19071 150 8 ) ) PUNCT bracis-19071 150 9 . . PUNCT bracis-19071 151 1 https://doi.org/10.1111/j.1469-1809.1936.tb02137.x https://doi.org/10.1111/j.1469-1809.1936.tb02137.x PROPN bracis-19071 151 2 , , PUNCT bracis-19071 151 3 https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1469-1809.1936.tb02137.x https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1469-1809.1936.tb02137.x PROPN bracis-19071 151 4 goodfellow goodfellow PROPN bracis-19071 151 5 , , PUNCT bracis-19071 151 6 i. i. PROPN bracis-19071 151 7 , , PUNCT bracis-19071 151 8 et et PROPN bracis-19071 151 9 al al PROPN bracis-19071 151 10 . . PUNCT bracis-19071 151 11 : : PUNCT bracis-19071 152 1 tensorflow tensorflow NOUN bracis-19071 152 2 : : PUNCT bracis-19071 152 3 large large ADJ bracis-19071 152 4 - - PUNCT bracis-19071 152 5 scale scale NOUN bracis-19071 152 6 machine machine NOUN bracis-19071 152 7 learning learn VERB bracis-19071 152 8 on on ADP bracis-19071 152 9 heterogeneous heterogeneous ADJ bracis-19071 152 10 systems system NOUN bracis-19071 152 11 ( ( PUNCT bracis-19071 152 12 2015 2015 NUM bracis-19071 152 13 ) ) PUNCT bracis-19071 152 14 . . PUNCT bracis-19071 153 1 https://www.tensorflow.org/software https://www.tensorflow.org/software PROPN bracis-19071 153 2 available available ADJ bracis-19071 153 3 from from ADP bracis-19071 153 4 tensorflow.org tensorflow.org X bracis-19071 153 5 he he PRON bracis-19071 153 6 , , PUNCT bracis-19071 153 7 k. k. PROPN bracis-19071 153 8 , , PUNCT bracis-19071 153 9 zhang zhang PROPN bracis-19071 153 10 , , PUNCT bracis-19071 153 11 x. x. PROPN bracis-19071 153 12 , , PUNCT bracis-19071 153 13 ren ren PROPN bracis-19071 153 14 , , PUNCT bracis-19071 153 15 s. s. PROPN bracis-19071 153 16 , , PUNCT bracis-19071 153 17 sun sun PROPN bracis-19071 153 18 , , PUNCT bracis-19071 153 19 j. j. PROPN bracis-19071 153 20 : : PUNCT bracis-19071 153 21 deep deep ADJ bracis-19071 153 22 residual residual ADJ bracis-19071 153 23 learning learning NOUN bracis-19071 153 24 for for ADP bracis-19071 153 25 image image NOUN bracis-19071 153 26 recognition recognition NOUN bracis-19071 153 27 . . PUNCT bracis-19071 154 1 in in ADP bracis-19071 154 2 : : PUNCT bracis-19071 154 3 proceedings proceeding NOUN bracis-19071 154 4 of of ADP bracis-19071 154 5 the the DET bracis-19071 154 6 ieee ieee NOUN bracis-19071 154 7 conference conference NOUN bracis-19071 154 8 on on ADP bracis-19071 154 9 computer computer NOUN bracis-19071 154 10 vision vision NOUN bracis-19071 154 11 and and CCONJ bracis-19071 154 12 pattern pattern NOUN bracis-19071 154 13 recognition recognition NOUN bracis-19071 154 14 , , PUNCT bracis-19071 154 15 pp pp ADP bracis-19071 154 16 . . PUNCT bracis-19071 155 1 770–778 770–778 NUM bracis-19071 155 2 ( ( PUNCT bracis-19071 155 3 2016 2016 NUM bracis-19071 155 4 ) ) PUNCT bracis-19071 155 5 google google PROPN bracis-19071 155 6 scholar scholar NOUN bracis-19071 155 7     SPACE bracis-19071 155 8 heredia heredia PROPN bracis-19071 155 9 , , PUNCT bracis-19071 155 10 i. i. NOUN bracis-19071 155 11 : : PUNCT bracis-19071 155 12 large large ADJ bracis-19071 155 13 - - PUNCT bracis-19071 155 14 scale scale NOUN bracis-19071 155 15 plant plant NOUN bracis-19071 155 16 classification classification NOUN bracis-19071 155 17 with with ADP bracis-19071 155 18 deep deep ADJ bracis-19071 155 19 neural neural ADJ bracis-19071 155 20 networks network NOUN bracis-19071 155 21 . . PUNCT bracis-19071 156 1 in in ADP bracis-19071 156 2 : : PUNCT bracis-19071 156 3 proceedings proceeding NOUN bracis-19071 156 4 of of ADP bracis-19071 156 5 the the DET bracis-19071 156 6 computing compute VERB bracis-19071 156 7 frontiers frontier NOUN bracis-19071 156 8 conference conference NOUN bracis-19071 156 9 , , PUNCT bracis-19071 156 10 pp pp X bracis-19071 156 11 . . PUNCT bracis-19071 157 1 259–262 259–262 NUM bracis-19071 157 2 ( ( PUNCT bracis-19071 157 3 2017 2017 NUM bracis-19071 157 4 ) ) PUNCT bracis-19071 157 5 google google PROPN bracis-19071 157 6 scholar scholar NOUN bracis-19071 157 7     SPACE bracis-19071 157 8 huang huang PROPN bracis-19071 157 9 , , PUNCT bracis-19071 157 10 g. g. PROPN bracis-19071 157 11 , , PUNCT bracis-19071 157 12 liu liu PROPN bracis-19071 157 13 , , PUNCT bracis-19071 157 14 z. z. PROPN bracis-19071 157 15 , , PUNCT bracis-19071 157 16 van van PROPN bracis-19071 157 17 der der NOUN bracis-19071 157 18 maaten maaten VERB bracis-19071 157 19 , , PUNCT bracis-19071 157 20 l. l. PROPN bracis-19071 157 21 , , PUNCT bracis-19071 157 22 weinberger weinberger PROPN bracis-19071 157 23 , , PUNCT bracis-19071 157 24 k.q k.q PROPN bracis-19071 157 25 . . PROPN bracis-19071 157 26 : : PUNCT bracis-19071 157 27 densely densely ADV bracis-19071 157 28 connected connect VERB bracis-19071 157 29 convolutional convolutional ADJ bracis-19071 157 30 networks network NOUN bracis-19071 157 31 . . PUNCT bracis-19071 158 1 in in ADP bracis-19071 158 2 : : PUNCT bracis-19071 158 3 proceedings proceeding NOUN bracis-19071 158 4 of of ADP bracis-19071 158 5 the the DET bracis-19071 158 6 ieee ieee NOUN bracis-19071 158 7 conference conference NOUN bracis-19071 158 8 on on ADP bracis-19071 158 9 computer computer NOUN bracis-19071 158 10 vision vision NOUN bracis-19071 158 11 and and CCONJ bracis-19071 158 12 pattern pattern NOUN bracis-19071 158 13 recognition recognition NOUN bracis-19071 158 14 , , PUNCT bracis-19071 158 15 pp pp X bracis-19071 158 16 . . PUNCT bracis-19071 159 1 4700–4708 4700–4708 NUM bracis-19071 159 2 ( ( PUNCT bracis-19071 159 3 2017 2017 NUM bracis-19071 159 4 ) ) PUNCT bracis-19071 159 5 google google PROPN bracis-19071 159 6 scholar scholar NOUN bracis-19071 159 7     SPACE bracis-19071 159 8 lecun lecun ADJ bracis-19071 159 9 , , PUNCT bracis-19071 159 10 y. y. PROPN bracis-19071 159 11 , , PUNCT bracis-19071 159 12 bottou bottou PROPN bracis-19071 159 13 , , PUNCT bracis-19071 159 14 l. l. PROPN bracis-19071 159 15 , , PUNCT bracis-19071 159 16 bengio bengio PROPN bracis-19071 159 17 , , PUNCT bracis-19071 159 18 y. y. PROPN bracis-19071 159 19 , , PUNCT bracis-19071 159 20 haffner haffner NOUN bracis-19071 159 21 , , PUNCT bracis-19071 159 22 p. p. NOUN bracis-19071 159 23 : : PUNCT bracis-19071 160 1 gradient gradient NOUN bracis-19071 160 2 - - PUNCT bracis-19071 160 3 based base VERB bracis-19071 160 4 learning learning NOUN bracis-19071 160 5 applied apply VERB bracis-19071 160 6 to to ADP bracis-19071 160 7 document document NOUN bracis-19071 160 8 recognition recognition NOUN bracis-19071 160 9 . . PUNCT bracis-19071 161 1 proc proc PROPN bracis-19071 161 2 . . PUNCT bracis-19071 162 1 ieee ieee PROPN bracis-19071 162 2 86(11 86(11 NUM bracis-19071 162 3 ) ) PUNCT bracis-19071 162 4 , , PUNCT bracis-19071 162 5 2278–2324 2278–2324 NUM bracis-19071 162 6 ( ( PUNCT bracis-19071 162 7 1998 1998 NUM bracis-19071 162 8 ) ) PUNCT bracis-19071 162 9 article article NOUN bracis-19071 162 10     SPACE bracis-19071 162 11 google google PROPN bracis-19071 162 12 scholar scholar NOUN bracis-19071 162 13     SPACE bracis-19071 162 14 loarie loarie NOUN bracis-19071 162 15 , , PUNCT bracis-19071 162 16 s. s. PROPN bracis-19071 162 17 : : PUNCT bracis-19071 162 18 a a DET bracis-19071 162 19 community community NOUN bracis-19071 162 20 for for ADP bracis-19071 162 21 naturalists naturalist NOUN bracis-19071 162 22 ( ( PUNCT bracis-19071 162 23 2008 2008 NUM bracis-19071 162 24 ) ) PUNCT bracis-19071 162 25 . . PUNCT bracis-19071 163 1 https://www.inaturalist.org/ https://www.inaturalist.org/ PROPN bracis-19071 163 2 pedregosa pedregosa PROPN bracis-19071 163 3 , , PUNCT bracis-19071 163 4 f. f. PROPN bracis-19071 163 5 , , PUNCT bracis-19071 163 6 et et PROPN bracis-19071 163 7 al al PROPN bracis-19071 163 8 . . PROPN bracis-19071 163 9 : : PUNCT bracis-19071 164 1 scikit scikit NOUN bracis-19071 164 2 - - PUNCT bracis-19071 164 3 learn learn VERB bracis-19071 164 4 : : PUNCT bracis-19071 164 5 machine machine NOUN bracis-19071 164 6 learning learning NOUN bracis-19071 164 7 in in ADP bracis-19071 164 8 python python PROPN bracis-19071 164 9 . . PUNCT bracis-19071 165 1 j. j. PROPN bracis-19071 165 2 mach mach PROPN bracis-19071 165 3 . . PUNCT bracis-19071 166 1 learn learn VERB bracis-19071 166 2 . . PUNCT bracis-19071 167 1 res re NOUN bracis-19071 167 2 . . PROPN bracis-19071 168 1 12 12 NUM bracis-19071 168 2 , , PUNCT bracis-19071 168 3 2825–2830 2825–2830 NUM bracis-19071 168 4 ( ( PUNCT bracis-19071 168 5 2011 2011 NUM bracis-19071 168 6 ) ) PUNCT bracis-19071 168 7 mathscinet mathscinet NOUN bracis-19071 168 8     SPACE bracis-19071 168 9 math math NOUN bracis-19071 168 10     SPACE bracis-19071 168 11 google google PROPN bracis-19071 168 12 scholar scholar NOUN bracis-19071 168 13     SPACE bracis-19071 168 14 r. r. PROPN bracis-19071 168 15 al al PROPN bracis-19071 168 16 - - PUNCT bracis-19071 168 17 qurran qurran PROPN bracis-19071 168 18 , , PUNCT bracis-19071 168 19 m.a.a m.a.a NOUN bracis-19071 168 20 . . PUNCT bracis-19071 168 21 , , PUNCT bracis-19071 168 22 shatnawi shatnawi PROPN bracis-19071 168 23 , , PUNCT bracis-19071 168 24 a. a. NOUN bracis-19071 168 25 : : PUNCT bracis-19071 168 26 plant plant NOUN bracis-19071 168 27 classification classification NOUN bracis-19071 168 28 in in ADP bracis-19071 168 29 the the DET bracis-19071 168 30 wild wild NOUN bracis-19071 168 31 : : PUNCT bracis-19071 168 32 a a DET bracis-19071 168 33 transfer transfer NOUN bracis-19071 168 34 learning learning NOUN bracis-19071 168 35 approach approach NOUN bracis-19071 168 36 . . PUNCT bracis-19071 169 1 international international ADJ bracis-19071 169 2 arab arab ADJ bracis-19071 169 3 conference conference PROPN bracis-19071 169 4 on on ADP bracis-19071 169 5 information information NOUN bracis-19071 169 6 technology technology NOUN bracis-19071 169 7 ( ( PUNCT bracis-19071 169 8 acit acit ADJ bracis-19071 169 9 ) ) PUNCT bracis-19071 169 10 ( ( PUNCT bracis-19071 169 11 2018 2018 NUM bracis-19071 169 12 ) ) PUNCT bracis-19071 169 13 google google PROPN bracis-19071 169 14 scholar scholar NOUN bracis-19071 169 15     SPACE bracis-19071 169 16 sandler sandler NOUN bracis-19071 169 17 , , PUNCT bracis-19071 169 18 m. m. NOUN bracis-19071 169 19 , , PUNCT bracis-19071 169 20 howard howard PROPN bracis-19071 169 21 , , PUNCT bracis-19071 169 22 a. a. PROPN bracis-19071 169 23 , , PUNCT bracis-19071 169 24 zhu zhu PROPN bracis-19071 169 25 , , PUNCT bracis-19071 169 26 m. m. NOUN bracis-19071 169 27 , , PUNCT bracis-19071 169 28 zhmoginov zhmoginov PROPN bracis-19071 169 29 , , PUNCT bracis-19071 169 30 a. a. PROPN bracis-19071 169 31 , , PUNCT bracis-19071 169 32 chen chen PROPN bracis-19071 169 33 , , PUNCT bracis-19071 169 34 l.c l.c PROPN bracis-19071 169 35 . . PROPN bracis-19071 169 36 : : PUNCT bracis-19071 169 37 mobilenetv mobilenetv ADJ bracis-19071 169 38 2 2 NUM bracis-19071 169 39 : : PUNCT bracis-19071 169 40 inverted inverted ADJ bracis-19071 169 41 residuals residual NOUN bracis-19071 169 42 and and CCONJ bracis-19071 169 43 linear linear ADJ bracis-19071 169 44 bottlenecks bottleneck NOUN bracis-19071 169 45 . . PUNCT bracis-19071 170 1 in in ADP bracis-19071 170 2 : : PUNCT bracis-19071 170 3 proceedings proceeding NOUN bracis-19071 170 4 of of ADP bracis-19071 170 5 the the DET bracis-19071 170 6 ieee ieee NOUN bracis-19071 170 7 conference conference NOUN bracis-19071 170 8 on on ADP bracis-19071 170 9 computer computer NOUN bracis-19071 170 10 vision vision NOUN bracis-19071 170 11 and and CCONJ bracis-19071 170 12 pattern pattern NOUN bracis-19071 170 13 recognition recognition NOUN bracis-19071 170 14 , , PUNCT bracis-19071 170 15 pp pp ADP bracis-19071 170 16 . . PUNCT bracis-19071 171 1 4510–4520 4510–4520 NUM bracis-19071 171 2 ( ( PUNCT bracis-19071 171 3 2018 2018 NUM bracis-19071 171 4 ) ) PUNCT bracis-19071 171 5 google google PROPN bracis-19071 171 6 scholar scholar NOUN bracis-19071 171 7     SPACE bracis-19071 171 8 szegedy szegedy PROPN bracis-19071 171 9 , , PUNCT bracis-19071 171 10 c. c. NOUN bracis-19071 171 11 , , PUNCT bracis-19071 171 12 vanhoucke vanhoucke PROPN bracis-19071 171 13 , , PUNCT bracis-19071 171 14 v. v. PROPN bracis-19071 171 15 , , PUNCT bracis-19071 171 16 ioffe ioffe PROPN bracis-19071 171 17 , , PUNCT bracis-19071 171 18 s. s. PROPN bracis-19071 171 19 , , PUNCT bracis-19071 171 20 shlens shlens PROPN bracis-19071 171 21 , , PUNCT bracis-19071 171 22 j. j. PROPN bracis-19071 171 23 , , PUNCT bracis-19071 171 24 wojna wojna NOUN bracis-19071 171 25 , , PUNCT bracis-19071 171 26 z. z. PROPN bracis-19071 171 27 : : PUNCT bracis-19071 171 28 rethinking rethink VERB bracis-19071 171 29 the the DET bracis-19071 171 30 inception inception ADJ bracis-19071 171 31 architecture architecture NOUN bracis-19071 171 32 for for ADP bracis-19071 171 33 computer computer NOUN bracis-19071 171 34 vision vision NOUN bracis-19071 171 35 . . PUNCT bracis-19071 172 1 in in ADP bracis-19071 172 2 : : PUNCT bracis-19071 172 3 proceedings proceeding NOUN bracis-19071 172 4 of of ADP bracis-19071 172 5 the the DET bracis-19071 172 6 ieee ieee NOUN bracis-19071 172 7 conference conference NOUN bracis-19071 172 8 on on ADP bracis-19071 172 9 computer computer NOUN bracis-19071 172 10 vision vision NOUN bracis-19071 172 11 and and CCONJ bracis-19071 172 12 pattern pattern NOUN bracis-19071 172 13 recognition recognition NOUN bracis-19071 172 14 , , PUNCT bracis-19071 172 15 pp pp ADP bracis-19071 172 16 . . PUNCT bracis-19071 173 1 2818–2826 2818–2826 NUM bracis-19071 173 2 ( ( PUNCT bracis-19071 173 3 2016 2016 NUM bracis-19071 173 4 ) ) PUNCT bracis-19071 173 5 google google PROPN bracis-19071 173 6 scholar scholar NOUN bracis-19071 173 7     SPACE bracis-19071 173 8 tan tan PROPN bracis-19071 173 9 , , PUNCT bracis-19071 173 10 m. m. NOUN bracis-19071 173 11 , , PUNCT bracis-19071 173 12 le le X bracis-19071 173 13 , , PUNCT bracis-19071 173 14 q. q. PROPN bracis-19071 173 15 : : PUNCT bracis-19071 173 16 efficientnet efficientnet PROPN bracis-19071 173 17 : : PUNCT bracis-19071 173 18 rethinking rethink VERB bracis-19071 173 19 model model NOUN bracis-19071 173 20 scaling scale VERB bracis-19071 173 21 for for ADP bracis-19071 173 22 convolutional convolutional ADJ bracis-19071 173 23 neural neural ADJ bracis-19071 173 24 networks network NOUN bracis-19071 173 25 . . PUNCT bracis-19071 174 1 in in ADP bracis-19071 174 2 : : PUNCT bracis-19071 174 3 international international ADJ bracis-19071 174 4 conference conference NOUN bracis-19071 174 5 on on ADP bracis-19071 174 6 machine machine NOUN bracis-19071 174 7 learning learning NOUN bracis-19071 174 8 , , PUNCT bracis-19071 174 9 pp pp ADP bracis-19071 174 10 . . PUNCT bracis-19071 174 11 6105–6114 6105–6114 NOUN bracis-19071 174 12 . . PUNCT bracis-19071 175 1 pmlr pmlr NOUN bracis-19071 175 2 ( ( PUNCT bracis-19071 175 3 2019 2019 NUM bracis-19071 175 4 ) ) PUNCT bracis-19071 175 5 google google PROPN bracis-19071 175 6 scholar scholar NOUN bracis-19071 175 7     SPACE bracis-19071 175 8 van van PROPN bracis-19071 175 9 horn horn NOUN bracis-19071 175 10 , , PUNCT bracis-19071 175 11 g. g. PROPN bracis-19071 175 12 , , PUNCT bracis-19071 175 13 et et PROPN bracis-19071 175 14 al al PROPN bracis-19071 175 15 . . PROPN bracis-19071 175 16 : : PUNCT bracis-19071 176 1 the the DET bracis-19071 176 2 inaturalist inaturalist ADJ bracis-19071 176 3 species specie VERB bracis-19071 176 4 classification classification NOUN bracis-19071 176 5 and and CCONJ bracis-19071 176 6 detection detection NOUN bracis-19071 176 7 dataset dataset NOUN bracis-19071 176 8 . . PUNCT bracis-19071 177 1 in in ADP bracis-19071 177 2 : : PUNCT bracis-19071 177 3 proceedings proceeding NOUN bracis-19071 177 4 of of ADP bracis-19071 177 5 the the DET bracis-19071 177 6 ieee ieee NOUN bracis-19071 177 7 conference conference NOUN bracis-19071 177 8 on on ADP bracis-19071 177 9 computer computer NOUN bracis-19071 177 10 vision vision NOUN bracis-19071 177 11 and and CCONJ bracis-19071 177 12 pattern pattern NOUN bracis-19071 177 13 recognition recognition NOUN bracis-19071 177 14 , , PUNCT bracis-19071 177 15 pp pp PROPN bracis-19071 177 16 . . PUNCT bracis-19071 178 1 8769–8778 8769–8778 NUM bracis-19071 178 2 ( ( PUNCT bracis-19071 178 3 2018 2018 NUM bracis-19071 178 4 ) ) PUNCT bracis-19071 178 5 google google PROPN bracis-19071 178 6 scholar scholar NOUN bracis-19071 178 7     SPACE bracis-19071 178 8 vanschoren vanschoren NOUN bracis-19071 178 9 , , PUNCT bracis-19071 178 10 j. j. PROPN bracis-19071 178 11 , , PUNCT bracis-19071 178 12 van van PROPN bracis-19071 178 13 rijn rijn PROPN bracis-19071 178 14 , , PUNCT bracis-19071 178 15 j.n j.n PROPN bracis-19071 178 16 . . PROPN bracis-19071 178 17 , , PUNCT bracis-19071 178 18 bischl bischl PROPN bracis-19071 178 19 , , PUNCT bracis-19071 178 20 b. b. PROPN bracis-19071 178 21 , , PUNCT bracis-19071 178 22 torgo torgo PROPN bracis-19071 178 23 , , PUNCT bracis-19071 178 24 l. l. PROPN bracis-19071 178 25 : : PUNCT bracis-19071 178 26 openml openml NOUN bracis-19071 178 27 : : PUNCT bracis-19071 178 28 networked network VERB bracis-19071 178 29 science science NOUN bracis-19071 178 30 in in ADP bracis-19071 178 31 machine machine NOUN bracis-19071 178 32 learning learning NOUN bracis-19071 178 33 . . PUNCT bracis-19071 179 1 sigkdd sigkdd VERB bracis-19071 179 2 explor explor NOUN bracis-19071 179 3 . . PUNCT bracis-19071 180 1 15(2 15(2 NUM bracis-19071 180 2 ) ) PUNCT bracis-19071 180 3 , , PUNCT bracis-19071 180 4 49–60 49–60 NUM bracis-19071 180 5 ( ( PUNCT bracis-19071 180 6 2013 2013 NUM bracis-19071 180 7 ) ) PUNCT bracis-19071 180 8 . . PUNCT bracis-19071 181 1 https://doi.org/10.1145/2641190.2641198 https://doi.org/10.1145/2641190.2641198 ADJ bracis-19071 181 2 , , PUNCT bracis-19071 181 3 http://doi.acm.org/10.1145/2641190.2641198 http://doi.acm.org/10.1145/2641190.2641198 PROPN bracis-19071 181 4 xiao xiao PROPN bracis-19071 181 5 , , PUNCT bracis-19071 181 6 h. h. PROPN bracis-19071 181 7 , , PUNCT bracis-19071 181 8 rasul rasul PROPN bracis-19071 181 9 , , PUNCT bracis-19071 181 10 k. k. PROPN bracis-19071 181 11 , , PUNCT bracis-19071 181 12 vollgraf vollgraf PROPN bracis-19071 181 13 , , PUNCT bracis-19071 181 14 r. r. PROPN bracis-19071 181 15 : : PUNCT bracis-19071 181 16 fashion fashion NOUN bracis-19071 181 17 - - PUNCT bracis-19071 181 18 mnist mnist NOUN bracis-19071 181 19 : : PUNCT bracis-19071 181 20 a a DET bracis-19071 181 21 novel novel ADJ bracis-19071 181 22 image image NOUN bracis-19071 181 23 dataset dataset VERB bracis-19071 181 24 for for ADP bracis-19071 181 25 benchmarking benchmarke VERB bracis-19071 181 26 machine machine NOUN bracis-19071 181 27 learning learn VERB bracis-19071 181 28 algorithms algorithm NOUN bracis-19071 181 29 ( ( PUNCT bracis-19071 181 30 2017 2017 NUM bracis-19071 181 31 ) ) PUNCT bracis-19071 181 32 . . PUNCT bracis-19071 182 1 arxiv arxiv PROPN bracis-19071 182 2 preprint preprint NOUN bracis-19071 182 3 arxiv:1708.07747 arxiv:1708.07747 PROPN bracis-19071 182 4 download download NOUN bracis-19071 182 5 references reference NOUN bracis-19071 182 6 author author NOUN bracis-19071 182 7 information information NOUN bracis-19071 182 8 authors author NOUN bracis-19071 182 9 and and CCONJ bracis-19071 182 10 affiliations affiliation NOUN bracis-19071 182 11 universidade universidade PROPN bracis-19071 182 12 federal federal PROPN bracis-19071 182 13 da da PROPN bracis-19071 182 14 paraíba paraíba PROPN bracis-19071 182 15 , , PUNCT bracis-19071 182 16 joão joão PROPN bracis-19071 182 17 pessoa pessoa PROPN bracis-19071 182 18 , , PUNCT bracis-19071 182 19 pb pb PROPN bracis-19071 182 20 , , PUNCT bracis-19071 182 21 brazil brazil PROPN bracis-19071 182 22 itamar itamar PROPN bracis-19071 182 23 de de PROPN bracis-19071 182 24 paiva paiva PROPN bracis-19071 182 25 rocha rocha PROPN bracis-19071 182 26 filho filho PROPN bracis-19071 182 27 , , PUNCT bracis-19071 182 28     SPACE bracis-19071 182 29 joão joão PROPN bracis-19071 182 30 pedro pedro PROPN bracis-19071 182 31 vasconcelos vasconcelos PROPN bracis-19071 182 32 teixeira teixeira PROPN bracis-19071 182 33 , , PUNCT bracis-19071 182 34     SPACE bracis-19071 182 35 joão joão PROPN bracis-19071 182 36 wallace wallace PROPN bracis-19071 182 37 lucena lucena PROPN bracis-19071 182 38 lins lins PROPN bracis-19071 182 39 , , PUNCT bracis-19071 182 40     SPACE bracis-19071 182 41 felipe felipe PROPN bracis-19071 182 42 honorato honorato PROPN bracis-19071 182 43 de de PROPN bracis-19071 182 44 sousa sousa PROPN bracis-19071 182 45 , , PUNCT bracis-19071 182 46     SPACE bracis-19071 182 47 ana ana PROPN bracis-19071 182 48 clara clara PROPN bracis-19071 182 49 chaves chaves PROPN bracis-19071 182 50 sousa sousa PROPN bracis-19071 182 51 , , PUNCT bracis-19071 182 52     SPACE bracis-19071 182 53 manuel manuel PROPN bracis-19071 182 54 ferreira ferreira PROPN bracis-19071 182 55 junior junior PROPN bracis-19071 182 56 , , PUNCT bracis-19071 182 57     SPACE bracis-19071 182 58 thaís thaís NOUN bracis-19071 182 59 gaudencio gaudencio NOUN bracis-19071 182 60 do do VERB bracis-19071 182 61 rêgo rêgo NOUN bracis-19071 182 62 , , PUNCT bracis-19071 182 63     SPACE bracis-19071 182 64 yuri yuri PROPN bracis-19071 182 65 de de PROPN bracis-19071 182 66 almeida almeida PROPN bracis-19071 182 67 malheiros malheiros PROPN bracis-19071 182 68     SPACE bracis-19071 182 69 & & CCONJ bracis-19071 182 70     SPACE bracis-19071 182 71 telmo telmo PROPN bracis-19071 182 72 silva silva PROPN bracis-19071 182 73 filho filho PROPN bracis-19071 182 74 universidade universidade PROPN bracis-19071 182 75 federal federal PROPN bracis-19071 182 76 do do AUX bracis-19071 182 77 rio rio PROPN bracis-19071 182 78 grande grande PROPN bracis-19071 182 79 do do AUX bracis-19071 182 80 norte norte NOUN bracis-19071 182 81 , , PUNCT bracis-19071 182 82 natal natal NOUN bracis-19071 182 83 , , PUNCT bracis-19071 182 84 rn rn PROPN bracis-19071 182 85 , , PUNCT bracis-19071 182 86 brazil brazil PROPN bracis-19071 182 87 thaís thaís PROPN bracis-19071 182 88 ramos ramos PROPN bracis-19071 182 89 universidade universidade PROPN bracis-19071 182 90 federal federal PROPN bracis-19071 182 91 de de PROPN bracis-19071 182 92 pernambuco pernambuco PROPN bracis-19071 182 93 , , PUNCT bracis-19071 182 94 recife recife NOUN bracis-19071 182 95 , , PUNCT bracis-19071 182 96 pe pe PROPN bracis-19071 182 97 , , PUNCT bracis-19071 182 98 brazil brazil PROPN bracis-19071 182 99 cecília cecília PROPN bracis-19071 182 100 silva silva PROPN bracis-19071 182 101 authors author NOUN bracis-19071 182 102 itamar itamar PROPN bracis-19071 182 103 de de PROPN bracis-19071 182 104 paiva paiva PROPN bracis-19071 182 105 rocha rocha PROPN bracis-19071 182 106 filhoview filhoview PROPN bracis-19071 182 107 author author NOUN bracis-19071 182 108 publications publication NOUN bracis-19071 182 109 search search NOUN bracis-19071 182 110 author author NOUN bracis-19071 182 111 on on ADP bracis-19071 182 112 : : PUNCT bracis-19071 182 113 pubmed pubmed PROPN bracis-19071 182 114     SPACE bracis-19071 182 115 google google PROPN bracis-19071 182 116 scholar scholar NOUN bracis-19071 182 117 joão joão PROPN bracis-19071 182 118 pedro pedro PROPN bracis-19071 182 119 vasconcelos vasconcelos PROPN bracis-19071 182 120 teixeiraview teixeiraview PROPN bracis-19071 182 121 author author NOUN bracis-19071 182 122 publications publication NOUN bracis-19071 182 123 search search NOUN bracis-19071 182 124 author author NOUN bracis-19071 182 125 on on ADP bracis-19071 182 126 : : PUNCT bracis-19071 182 127 pubmed pubmed PROPN bracis-19071 182 128     SPACE bracis-19071 182 129 google google PROPN bracis-19071 182 130 scholar scholar PROPN bracis-19071 182 131 joão joão PROPN bracis-19071 182 132 wallace wallace PROPN bracis-19071 182 133 lucena lucena PROPN bracis-19071 182 134 linsview linsview PROPN bracis-19071 182 135 author author NOUN bracis-19071 182 136 publications publication NOUN bracis-19071 182 137 search search NOUN bracis-19071 182 138 author author NOUN bracis-19071 182 139 on on ADP bracis-19071 182 140 : : PUNCT bracis-19071 182 141 pubmed pubmed PROPN bracis-19071 182 142     SPACE bracis-19071 182 143 google google PROPN bracis-19071 182 144 scholar scholar NOUN bracis-19071 182 145 felipe felipe PROPN bracis-19071 182 146 honorato honorato PROPN bracis-19071 182 147 de de PROPN bracis-19071 182 148 sousaview sousaview NOUN bracis-19071 182 149 author author NOUN bracis-19071 182 150 publications publication NOUN bracis-19071 182 151 search search NOUN bracis-19071 182 152 author author NOUN bracis-19071 182 153 on on ADP bracis-19071 182 154 : : PUNCT bracis-19071 182 155 pubmed pubmed PROPN bracis-19071 182 156     SPACE bracis-19071 182 157 google google PROPN bracis-19071 182 158 scholar scholar PROPN bracis-19071 182 159 ana ana PROPN bracis-19071 182 160 clara clara PROPN bracis-19071 182 161 chaves chaves PROPN bracis-19071 182 162 sousaview sousaview PROPN bracis-19071 182 163 author author NOUN bracis-19071 182 164 publications publication NOUN bracis-19071 182 165 search search NOUN bracis-19071 182 166 author author NOUN bracis-19071 182 167 on on ADP bracis-19071 182 168 : : PUNCT bracis-19071 182 169 pubmed pubmed PROPN bracis-19071 182 170     SPACE bracis-19071 182 171 google google PROPN bracis-19071 182 172 scholar scholar NOUN bracis-19071 182 173 manuel manuel PROPN bracis-19071 182 174 ferreira ferreira PROPN bracis-19071 182 175 juniorview juniorview PROPN bracis-19071 182 176 author author NOUN bracis-19071 182 177 publications publication NOUN bracis-19071 182 178 search search NOUN bracis-19071 182 179 author author NOUN bracis-19071 182 180 on on ADP bracis-19071 182 181 : : PUNCT bracis-19071 182 182 pubmed pubmed PROPN bracis-19071 182 183     SPACE bracis-19071 182 184 google google PROPN bracis-19071 182 185 scholar scholar NOUN bracis-19071 182 186 thaís thaís NOUN bracis-19071 182 187 ramosview ramosview NOUN bracis-19071 182 188 author author NOUN bracis-19071 182 189 publications publication NOUN bracis-19071 182 190 search search NOUN bracis-19071 182 191 author author NOUN bracis-19071 182 192 on on ADP bracis-19071 182 193 : : PUNCT bracis-19071 182 194 pubmed pubmed PROPN bracis-19071 182 195     SPACE bracis-19071 182 196 google google PROPN bracis-19071 182 197 scholar scholar NOUN bracis-19071 182 198 cecília cecília NOUN bracis-19071 182 199 silvaview silvaview VERB bracis-19071 182 200 author author NOUN bracis-19071 182 201 publications publication NOUN bracis-19071 182 202 search search NOUN bracis-19071 182 203 author author NOUN bracis-19071 182 204 on on ADP bracis-19071 182 205 : : PUNCT bracis-19071 182 206 pubmed pubmed PROPN bracis-19071 182 207     SPACE bracis-19071 182 208 google google PROPN bracis-19071 182 209 scholar scholar NOUN bracis-19071 182 210 thaís thaís NOUN bracis-19071 182 211 gaudencio gaudencio NOUN bracis-19071 182 212 do do AUX bracis-19071 182 213 rêgoview rêgoview VERB bracis-19071 182 214 author author NOUN bracis-19071 182 215 publications publication NOUN bracis-19071 182 216 search search NOUN bracis-19071 182 217 author author NOUN bracis-19071 182 218 on on ADP bracis-19071 182 219 : : PUNCT bracis-19071 182 220 pubmed pubmed PROPN bracis-19071 182 221     SPACE bracis-19071 182 222 google google PROPN bracis-19071 182 223 scholar scholar NOUN bracis-19071 182 224 yuri yuri PROPN bracis-19071 182 225 de de PROPN bracis-19071 182 226 almeida almeida PROPN bracis-19071 182 227 malheirosview malheirosview PROPN bracis-19071 182 228 author author NOUN bracis-19071 182 229 publications publication NOUN bracis-19071 182 230 search search NOUN bracis-19071 182 231 author author NOUN bracis-19071 182 232 on on ADP bracis-19071 182 233 : : PUNCT bracis-19071 182 234 pubmed pubmed PROPN bracis-19071 182 235     SPACE bracis-19071 182 236 google google PROPN bracis-19071 182 237 scholar scholar NOUN bracis-19071 182 238 telmo telmo PROPN bracis-19071 182 239 silva silva PROPN bracis-19071 182 240 filhoview filhoview PROPN bracis-19071 182 241 author author NOUN bracis-19071 182 242 publications publication NOUN bracis-19071 182 243 search search NOUN bracis-19071 182 244 author author NOUN bracis-19071 182 245 on on ADP bracis-19071 182 246 : : PUNCT bracis-19071 182 247 pubmed pubmed PROPN bracis-19071 182 248     SPACE bracis-19071 182 249 google google PROPN bracis-19071 182 250 scholar scholar NOUN bracis-19071 182 251 editor editor NOUN bracis-19071 182 252 information information NOUN bracis-19071 182 253 editors editor NOUN bracis-19071 182 254 and and CCONJ bracis-19071 182 255 affiliations affiliation NOUN bracis-19071 182 256 universidade universidade PROPN bracis-19071 182 257 federal federal PROPN bracis-19071 182 258 de de X bracis-19071 182 259 sergipe sergipe PROPN bracis-19071 182 260 , , PUNCT bracis-19071 182 261 são são NOUN bracis-19071 182 262 cristóvão cristóvão PROPN bracis-19071 182 263 , , PUNCT bracis-19071 182 264 brazil brazil PROPN bracis-19071 182 265 andré andré PROPN bracis-19071 182 266 britto britto PROPN bracis-19071 182 267 universidade universidade PROPN bracis-19071 182 268 de de PROPN bracis-19071 182 269 são são PROPN bracis-19071 182 270 paulo paulo PROPN bracis-19071 182 271 , , PUNCT bracis-19071 182 272 são são PROPN bracis-19071 182 273 paulo paulo PROPN bracis-19071 182 274 , , PUNCT bracis-19071 182 275 brazil brazil PROPN bracis-19071 182 276 karina karina PROPN bracis-19071 182 277 valdivia valdivia PROPN bracis-19071 182 278 delgado delgado VERB bracis-19071 182 279 rights right NOUN bracis-19071 182 280 and and CCONJ bracis-19071 182 281 permissions permission NOUN bracis-19071 182 282 reprints reprint NOUN bracis-19071 182 283 and and CCONJ bracis-19071 182 284 permissions permission VERB bracis-19071 182 285 copyright copyright NOUN bracis-19071 182 286 information information NOUN bracis-19071 182 287 © © PROPN bracis-19071 182 288 2021 2021 NUM bracis-19071 182 289 springer springer NOUN bracis-19071 182 290 nature nature NOUN bracis-19071 182 291 switzerland switzerland PROPN bracis-19071 182 292 ag ag PROPN bracis-19071 182 293 about about ADP bracis-19071 182 294 this this DET bracis-19071 182 295 paper paper NOUN bracis-19071 182 296 cite cite VERB bracis-19071 182 297 this this DET bracis-19071 182 298 paper paper NOUN bracis-19071 182 299 de de PROPN bracis-19071 182 300 paiva paiva PROPN bracis-19071 182 301 rocha rocha PROPN bracis-19071 182 302 filho filho PROPN bracis-19071 182 303 , , PUNCT bracis-19071 182 304 i. i. PROPN bracis-19071 182 305 et et PROPN bracis-19071 182 306 al al PROPN bracis-19071 182 307 . . PROPN bracis-19071 183 1 ( ( PUNCT bracis-19071 183 2 2021 2021 NUM bracis-19071 183 3 ) ) PUNCT bracis-19071 183 4 . . PUNCT bracis-19071 184 1 iris iris PROPN bracis-19071 184 2 - - PUNCT bracis-19071 184 3 cv cv PROPN bracis-19071 184 4 : : PUNCT bracis-19071 184 5 classifying classify VERB bracis-19071 184 6 iris iris NOUN bracis-19071 184 7 flowers flower NOUN bracis-19071 184 8 is be AUX bracis-19071 184 9 not not PART bracis-19071 184 10 as as ADV bracis-19071 184 11 easy easy ADJ bracis-19071 184 12 as as SCONJ bracis-19071 184 13 you you PRON bracis-19071 184 14 thought think VERB bracis-19071 184 15 . . PUNCT bracis-19071 185 1 in in ADP bracis-19071 185 2 : : PUNCT bracis-19071 185 3 britto britto PROPN bracis-19071 185 4 , , PUNCT bracis-19071 185 5 a. a. PROPN bracis-19071 185 6 , , PUNCT bracis-19071 185 7 valdivia valdivia PROPN bracis-19071 185 8 delgado delgado PROPN bracis-19071 185 9 , , PUNCT bracis-19071 185 10 k. k. PROPN bracis-19071 185 11 ( ( PUNCT bracis-19071 185 12 eds eds PROPN bracis-19071 185 13 ) ) PUNCT bracis-19071 185 14 intelligent intelligent ADJ bracis-19071 185 15 systems system NOUN bracis-19071 185 16 . . PUNCT bracis-19071 186 1 bracis bracis PROPN bracis-19071 186 2 2021 2021 NUM bracis-19071 186 3 . . PUNCT bracis-19071 187 1 lecture lecture NOUN bracis-19071 187 2 notes note NOUN bracis-19071 187 3 in in ADP bracis-19071 187 4 computer computer NOUN bracis-19071 187 5 science science NOUN bracis-19071 187 6 ( ( PUNCT bracis-19071 187 7 ) ) PUNCT bracis-19071 187 8 , , PUNCT bracis-19071 187 9 vol vol NOUN bracis-19071 187 10 13074 13074 NUM bracis-19071 187 11 . . PUNCT bracis-19071 188 1 springer springer NOUN bracis-19071 188 2 , , PUNCT bracis-19071 188 3 cham cham PROPN bracis-19071 188 4 . . PUNCT bracis-19071 189 1 https://doi.org/10.1007/978-3-030-91699-2_18 https://doi.org/10.1007/978-3-030-91699-2_18 PROPN bracis-19071 189 2 download download NOUN bracis-19071 189 3 citation citation NOUN bracis-19071 189 4 .ris .ris PUNCT bracis-19071 189 5 .enw .enw PROPN bracis-19071 189 6 .bib .bib PUNCT bracis-19071 190 1 doi doi PROPN bracis-19071 190 2 : : PUNCT bracis-19071 190 3 https://doi.org/10.1007/978-3-030-91699-2_18 https://doi.org/10.1007/978-3-030-91699-2_18 PROPN bracis-19071 190 4 published publish VERB bracis-19071 190 5 : : PUNCT bracis-19071 190 6 28 28 NUM bracis-19071 190 7 november november PROPN bracis-19071 190 8 2021 2021 NUM bracis-19071 190 9 publisher publisher NOUN bracis-19071 190 10 name name NOUN bracis-19071 190 11 : : PUNCT bracis-19071 190 12 springer springer NOUN bracis-19071 190 13 , , PUNCT bracis-19071 190 14 cham cham PROPN bracis-19071 190 15 print print PROPN bracis-19071 190 16 isbn isbn PROPN bracis-19071 190 17 : : PUNCT bracis-19071 190 18 978 978 NUM bracis-19071 190 19 - - SYM bracis-19071 190 20 3 3 NUM bracis-19071 190 21 - - PUNCT bracis-19071 190 22 030 030 NUM bracis-19071 190 23 - - PUNCT bracis-19071 190 24 91698 91698 NUM bracis-19071 190 25 - - SYM bracis-19071 190 26 5 5 NUM bracis-19071 190 27 online online ADJ bracis-19071 190 28 isbn isbn NOUN bracis-19071 190 29 : : PUNCT bracis-19071 190 30 978 978 NUM bracis-19071 190 31 - - SYM bracis-19071 190 32 3 3 NUM bracis-19071 190 33 - - PUNCT bracis-19071 190 34 030 030 NUM bracis-19071 190 35 - - PUNCT bracis-19071 190 36 91699 91699 NUM bracis-19071 190 37 - - SYM bracis-19071 190 38 2 2 NUM bracis-19071 190 39 ebook ebook NOUN bracis-19071 190 40 packages package NOUN bracis-19071 190 41 : : PUNCT bracis-19071 190 42 computer computer NOUN bracis-19071 190 43 sciencecomputer sciencecomputer NOUN bracis-19071 190 44 science science NOUN bracis-19071 190 45 ( ( PUNCT bracis-19071 190 46 r0 r0 NOUN bracis-19071 190 47 ) ) PUNCT bracis-19071 190 48 share share VERB bracis-19071 190 49 this this DET bracis-19071 190 50 paper paper NOUN bracis-19071 190 51 anyone anyone PRON bracis-19071 190 52 you you PRON bracis-19071 190 53 share share VERB bracis-19071 190 54 the the DET bracis-19071 190 55 following follow VERB bracis-19071 190 56 link link NOUN bracis-19071 190 57 with with ADP bracis-19071 190 58 will will AUX bracis-19071 190 59 be be AUX bracis-19071 190 60 able able ADJ bracis-19071 190 61 to to PART bracis-19071 190 62 read read VERB bracis-19071 190 63 this this DET bracis-19071 190 64 content content NOUN bracis-19071 190 65 : : PUNCT bracis-19071 190 66 get get VERB bracis-19071 190 67 shareable shareable ADJ bracis-19071 190 68 linksorry linksorry NOUN bracis-19071 190 69 , , PUNCT bracis-19071 190 70 a a DET bracis-19071 190 71 shareable shareable ADJ bracis-19071 190 72 link link NOUN bracis-19071 190 73 is be AUX bracis-19071 190 74 not not PART bracis-19071 190 75 currently currently ADV bracis-19071 190 76 available available ADJ bracis-19071 190 77 for for ADP bracis-19071 190 78 this this DET bracis-19071 190 79 article article NOUN bracis-19071 190 80 . . PUNCT bracis-19071 191 1 copy copy VERB bracis-19071 191 2 shareable shareable ADJ bracis-19071 191 3 link link NOUN bracis-19071 191 4 to to PART bracis-19071 191 5 clipboard clipboard NOUN bracis-19071 191 6 provided provide VERB bracis-19071 191 7 by by ADP bracis-19071 191 8 the the DET bracis-19071 191 9 springer springer NOUN bracis-19071 191 10 nature nature PROPN bracis-19071 191 11 sharedit sharedit PROPN bracis-19071 191 12 content content NOUN bracis-19071 191 13 - - PUNCT bracis-19071 191 14 sharing share VERB bracis-19071 191 15 initiative initiative NOUN bracis-19071 191 16 keywords keyword NOUN bracis-19071 191 17 computer computer NOUN bracis-19071 191 18 vision vision NOUN bracis-19071 191 19 dataset dataset NOUN bracis-19071 191 20 machine machine NOUN bracis-19071 191 21 learning learning NOUN bracis-19071 191 22 publish publish VERB bracis-19071 191 23 with with ADP bracis-19071 191 24 us us PROPN bracis-19071 191 25 policies policy NOUN bracis-19071 191 26 and and CCONJ bracis-19071 191 27 ethics ethic NOUN bracis-19071 191 28 search search NOUN bracis-19071 191 29 search search NOUN bracis-19071 191 30 by by ADP bracis-19071 191 31 keyword keyword NOUN bracis-19071 191 32 or or CCONJ bracis-19071 191 33 author author NOUN bracis-19071 191 34 search search NOUN bracis-19071 191 35 navigation navigation NOUN bracis-19071 191 36 find find VERB bracis-19071 191 37 a a DET bracis-19071 191 38 journal journal NOUN bracis-19071 191 39 publish publish VERB bracis-19071 191 40 with with ADP bracis-19071 191 41 us we PRON bracis-19071 191 42 track track VERB bracis-19071 191 43 your your PRON bracis-19071 191 44 research research NOUN bracis-19071 191 45 discover discover VERB bracis-19071 191 46 content content NOUN bracis-19071 191 47 journals journal NOUN bracis-19071 191 48 a a DET bracis-19071 191 49 - - PUNCT bracis-19071 191 50 z z NOUN bracis-19071 191 51 books book NOUN bracis-19071 191 52 a a DET bracis-19071 191 53 - - PUNCT bracis-19071 191 54 z z NOUN bracis-19071 191 55 publish publish NOUN bracis-19071 191 56 with with ADP bracis-19071 191 57 us us PROPN bracis-19071 191 58 journal journal PROPN bracis-19071 191 59 finder finder PROPN bracis-19071 191 60 publish publish VERB bracis-19071 191 61 your your PRON bracis-19071 191 62 research research NOUN bracis-19071 191 63 language language NOUN bracis-19071 191 64 editing edit VERB bracis-19071 191 65 open open ADJ bracis-19071 191 66 access access NOUN bracis-19071 191 67 publishing publishing NOUN bracis-19071 191 68 products product NOUN bracis-19071 191 69 and and CCONJ bracis-19071 191 70 services service NOUN bracis-19071 191 71 our our PRON bracis-19071 191 72 products product NOUN bracis-19071 191 73 librarians librarian VERB bracis-19071 191 74 societies society NOUN bracis-19071 191 75 partners partner NOUN bracis-19071 191 76 and and CCONJ bracis-19071 191 77 advertisers advertiser NOUN bracis-19071 191 78 our our PRON bracis-19071 191 79 brands brand NOUN bracis-19071 191 80 springer springer NOUN bracis-19071 191 81 nature nature PROPN bracis-19071 191 82 portfolio portfolio PROPN bracis-19071 191 83 bmc bmc PROPN bracis-19071 191 84 palgrave palgrave PROPN bracis-19071 191 85 macmillan macmillan PROPN bracis-19071 191 86 apress apress PROPN bracis-19071 191 87 discover discover VERB bracis-19071 191 88 your your PRON bracis-19071 191 89 privacy privacy NOUN bracis-19071 191 90 choices choice NOUN bracis-19071 191 91 / / SYM bracis-19071 191 92 manage manage NOUN bracis-19071 191 93 cookies cookie NOUN bracis-19071 191 94 your your PRON bracis-19071 191 95 us us PROPN bracis-19071 192 1 state state NOUN bracis-19071 192 2 privacy privacy NOUN bracis-19071 192 3 rights right NOUN bracis-19071 192 4 accessibility accessibility NOUN bracis-19071 192 5 statement statement NOUN bracis-19071 192 6 terms term NOUN bracis-19071 192 7 and and CCONJ bracis-19071 192 8 conditions condition NOUN bracis-19071 192 9 privacy privacy NOUN bracis-19071 192 10 policy policy NOUN bracis-19071 192 11 help help NOUN bracis-19071 192 12 and and CCONJ bracis-19071 192 13 support support VERB bracis-19071 192 14 legal legal ADJ bracis-19071 192 15 notice notice NOUN bracis-19071 192 16 cancel cancel VERB bracis-19071 192 17 contracts contract NOUN bracis-19071 192 18 here here ADV bracis-19071 192 19 129.74.145.123 129.74.145.123 NUM bracis-19071 192 20 hesburgh hesburgh PROPN bracis-19071 192 21 library library PROPN bracis-19071 192 22 er er INTJ bracis-19071 192 23 unit unit NOUN bracis-19071 192 24 ( ( PUNCT bracis-19071 192 25 3005732405 3005732405 NUM bracis-19071 192 26 ) ) PUNCT bracis-19071 192 27 northeast northeast ADJ bracis-19071 192 28 research research NOUN bracis-19071 192 29 libraries library NOUN bracis-19071 192 30 ( ( PUNCT bracis-19071 192 31 nerl nerl PROPN bracis-19071 192 32 ) ) PUNCT bracis-19071 192 33 ( ( PUNCT bracis-19071 192 34 8200828607 8200828607 NUM bracis-19071 192 35 ) ) PUNCT bracis-19071 192 36 nerl nerl VERB bracis-19071 192 37 ta ta X bracis-19071 192 38 account account NOUN bracis-19071 192 39 ( ( PUNCT bracis-19071 192 40 3006206169 3006206169 NUM bracis-19071 192 41 ) ) PUNCT bracis-19071 192 42 university university NOUN bracis-19071 192 43 of of ADP bracis-19071 192 44 notre notre PROPN bracis-19071 192 45 dame dame PROPN bracis-19071 192 46 hesburgh hesburgh PROPN bracis-19071 192 47 library library NOUN bracis-19071 192 48 ( ( PUNCT bracis-19071 192 49 3000184373 3000184373 NUM bracis-19071 192 50 ) ) PUNCT bracis-19071 193 1 © © ADP bracis-19071 193 2 2025 2025 NUM bracis-19071 193 3 springer springer NOUN bracis-19071 193 4 nature nature NOUN