id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
ajst-905	Zhu, Chenhao; Ye, Xia; Lu, Qiduo	Feature-Fusion Parallel Decoding Transformer for Image Captioning	2022	7	.pdf	application/pdf	3500	264	66	Among them, the CNN and RNN are used as encoders and decoders to extract image features and generate corresponding image descriptions, respectively. [9] where the grid size is 7 × 7 and the dimension of image features is 2048.	cache/ajst-905.pdf	txt/ajst-905.txt