LLaMA-VID: An Image is Worth 2 Tokens in Large Language Models.

Yanwei Li, Chengyao Wang, Jiaya Jia

Browse the full ECCV paper archive.