ShadowDancer: Teaching Video World Models Any Action by Learning Unified Dynamics Representations from a Video and Its Shadow

  • 类型：arxiv
  • 标识：2607.28362
  • 链接：https://arxiv.org/abs/2607.28362
  • 主分类：multimodal
  • 形态：method
  • 被引：0
  • 被引来源：Semantic Scholar + OpenAlex
  • S2被引：0
  • OpenAlex被引：0
  • 影响力被引：0
  • TLDR：ShadowDancer introduces shadow pairs, video pairs that replay the same dynamics under independently resampled appearance, constructed at scale by the Shadow Library, so that a dynamics family becomes controllable exactly when such pairs can be constructed for it.
  • OpenAlex ID：W7171924906
  • OpenAlex DOI：10.48550/arxiv.2607.28362
  • DOI：10.48550/arxiv.2607.28362
  • DOI来源：OpenAlex
  • 开放获取：green
  • 开放获取链接：https://doi.org/10.48550/arxiv.2607.28362
  • OpenAlex更新：2026-08-26
  • 待LLM分类：否
  • 标题中文：ShadowDancer: 通过从视频及其阴影中学习统一动力学表示来教视频世界模型执行任意动作
  • TLDR中文：ShadowDancer 引入 shadow pair，即在同一动力学下对外观做独立重采样的成对视频，并由 Shadow Library 大规模构建；一个 dynamics family 可控，当且仅当能为其构造出这样的 pair。
  • 来源文件：
  • /inbox/tom/_candidates/2026-07-31-agent-rag-longcontext-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]