xGen-MM-Vid (BLIP-3-Video): You Only Need 32 Tokens to Represent a Video Even in VLMs
Research paper by Michael S. Ryoo and collaborators. Published 2024-10-21. Topics: multimodal language model, temporal information, temporal encoder, visual tokenizer, spatio-temporal pooling.
Resource information
- Authors
- Michael S. Ryoo, Honglu Zhou, Shrikant Kendre, Can Qin, Le Xue, Manli Shu, Silvio Savarese, Ran Xu, Caiming Xiong, Juan Carlos Niebles
- Research topics
- multimodal language model, temporal information, temporal encoder, visual tokenizer, spatio-temporal pooling, Token Turing Machines, video question-answering, visual tokens
- Publication date
- 2024-10-21
- Publication
- arXiv
- Last record update
- Nov 3, 2026
