arXiv Artificial Intelligence· Xinyu Chen, Song Wu, Dingbao Shao, Jiahang Li, Jiang Lin, Yuhang Liu, Duo Li, Jiaxin Hu, Shengrong Gu, Liang Li, Qian Wang, Junlan Feng, Ying Tai, Jian Yang, Zili Yi·· 17 小时前AI 评分40
SemantiScene-Edit：面向指令视频编辑的专家合成监督与双路径源条件
SemantiScene-Edit: Expert-Synthesized Supervision and Dual-Path Source Conditioning for Instruction-Based Video Editing
AI 导读
SemantiScene-Edit 将多模态编码器与预训练视频扩散 Transformer 结合，实现按指令编辑视频并保留源视频外观与时序结构。模型采用双路径设计：可学习查询经可训练连接器和交叉注意力传递语义上下文，视觉路径在 patch embedding 前拼接源视频 latent 与噪声目标 latent。
来源：arXiv Artificial Intelligence · arxiv.org