arXiv cs.AI· Xiao Cao, Fuchen Long, Yansong Qu, Wen Xiao, Xin Zhang, Xinlong Sun, Xi Chen, Zhiyong Huang·· 3 小时前AI 评分37
VideoEditor-MoE：用混合专家架构扩展复杂视频编辑
VideoEditor-MoE: Scaling Complex Video Editing with Mixture of Experts
AI 导读
首个开源视频编辑混合专家框架 VideoEditor-MoE 发布。它基于 5B 视频编辑模型初始化，扩展至 14B 和 56B 参数，每个 token 激活约 8B 参数，并引入 MoE-Attention 通过路由 key-value 专家扩展注意力容量。分析显示专家路由更多响应样本内容与扩散噪声水平而非人工定义的编辑类别，且解耦权重衰减可稳定专家扩展。
正文 · 原文
Abstract:Complex video editing requires understanding complex instructions while jointly fulfilling multiple editing requests. Meeting these demands through dense model scaling substantially increases computational cost. We introduce VideoEditor-MoE, the first open-source mixture-of-experts framework for video editing. Beyond sparse MLP-MoE layers, MoE-Attention expands attention capacity through routed key--value experts while preserving global interactions among video tokens. Initialized from a 5B video editing model, VideoEditor-MoE scales to 14B and 56B parameters, with approximately 8B parameters activated per token in both variants. Comparisons with existing methods and two complementary baselines, Explicit-MoE and MLP-MoE, assess its benefits over explicit expert guidance and MLP-only expert architectures. Further analyses demonstrate that all experts remain active and reveal that learned routing responds more strongly to sample content and diffusion noise levels than to human-defined editing categories. Spatial analysis further indicates sensitivity to object-level semantic features. We also identify decoupled weight decay as a strategy for stabilizing expert scaling.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.03043 [cs.CV] |
| (or arXiv:2611.03043v3 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.03043 arXiv-issued DOI via DataCite |
Submission history
From: Xiao Cao [view email]
[v1]
Mon, 2 Nov 2026 12:42:51 UTC (6,654 KB)
[v2]
Mon, 9 Nov 2026 13:11:07 UTC (6,658 KB)
[v3]
Wed, 11 Nov 2026 11:48:20 UTC (6,655 KB)
来源：arXiv cs.AI · arxiv.org