CNN-Encoder and RNN-Decoder (Bahdanau Attention) for image caption or image to text on MS-COCO dataset. 图片描述