Linear Scaling Video VLMs for Long Video Understanding 事件

Linear Scaling Video VLMs for Long Video Understanding arXiv:2605.31598v1 Announce Type: new Abstract: Video vision-language models (VLMs) are increasingly used in long-horizon and streaming settings, yet most video encoders still rely on spatiotemporal self-attention, causing compute and latency to grow quadratically with the number of frames. Existing efficiency methods improve scalability but often lose accuracy relative to full self-attention, for example through aggressive frame/token drop