M-RAT: a Multi-grained Retrieval Augmentation Transformer for Image Captioning.

Jiayan Song, Renjie Pan, Jun Zhou, Hua Yang

Browse the full ACCV paper archive.