https://speakerdeck.com/fztkm 2025-08-09 03:21:34 -0400 VLMのトークン削減論文まとめ ・Keyframe-oriented Vision Token Pruning: Enhancing Efficiency of Large Vision Language Models on Long-Form Video Processing ・Multi-Granular Spatio-Temporal Token Merging for Training-Free Acceleration of Video LLMs ・METEOR: Multi-Encoder Collaborative Token Pruning for Efficient Vision Language Models VLMのトークン削減論文まとめ ・Keyframe-oriented Vision Token Pruning: Enhancing Efficiency of Large Vision Language Models on Long-Form Video Processing ・Multi-Granular Spatio-Temporal Token Merging for Training-Free Acceleration of Video LLMs ・METEOR: Multi-Encoder Collaborative Token Pruning for Efficient Vision Language Models Mon, 15 Dec 2025 00:00:00 -0500 https://speakerdeck.com/fztkm/iccv2025lun-wen-matome-vlmniokerutokunxue-jian-kvtp-sttm-meteor https://speakerdeck.com/fztkm/iccv2025lun-wen-matome-vlmniokerutokunxue-jian-kvtp-sttm-meteor Sat, 09 Aug 2025 00:00:00 -0400 https://speakerdeck.com/fztkm/2025-06-12-utd-visa-motionmae https://speakerdeck.com/fztkm/2025-06-12-utd-visa-motionmae