LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding

Devs

LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding | Read Paper on Bytez