xGen-MM-Vid (BLIP-3-Video): You Only Need 32 Tokens to Represent a Video Even in VLMs
Paper
•
2410.16267
•
Published
•
18
•
2
Michael Ryoo
michaelryoo
AI & ML interests
None yet
Organizations
commented
a
paper
over 1 year ago