1
Audio-Visual Exchange-Aware Token Pruning for Efficient Audio-Visual Captioning
新方法通过交换感知的令牌剪枝,高效提升音频-视觉字幕生成性能,论文来自arXiv最新研究。
arXiv:2606.10533v1 Announce Type: new Abstract: Audio-visual captioning generates natural language descriptions from video and audio content. Multimod…