Say in Human-Like Way: Hierarchical Cross-modal Information Abstraction and Summarization for Controllable Captioning.

Xiaoyi Wang, Jun Huang

Browse the full ICANN paper archive.