Improving Cross-Modal Alignment with Synthetic Pairs for Text-Only Image Captioning.

Zhiyue Liu, Jinyuan Liu, Fanrong Ma

Browse the full AAAI paper archive.