Efficient Object-Level Visual Context Modeling for Multimodal Machine Translation: Masking Irrelevant Objects Helps Grounding.

Dexin Wang, Deyi Xiong

Browse the full AAAI paper archive.