AnyTalk: Speech Animation for Arbitrary Characters Leveraging a Video Generation Model

  • 类型：arxiv
  • 标识：2608.16143
  • 链接：https://arxiv.org/abs/2608.16143
  • 主分类：multimodal
  • 形态：method
  • 被引：0
  • 被引来源：Semantic Scholar
  • S2被引：0
  • 影响力被引：0
  • TLDR：AnyTalk enables lip-synced animations across diverse face meshes and blendshape configurations, significantly reducing manual effort and data requirements and enhances usability by distilling AnyTalk into a streamlined network, $\text{AnyTalk}_{RT}$, thereby enabling real-time performance.
  • 副分类：engineering
  • 待LLM分类：否
  • 标题中文：AnyTalk：基于视频生成模型的任意角色语音动画
  • TLDR中文：AnyTalk 能够在多样化的人脸网格和 blendshape 配置下生成唇形同步动画，显著减少人工工作和数据需求；并通过将 AnyTalk 蒸馏为精简网络 $\text{AnyTalk}_{RT}$ 来提升可用性，从而实现实时性能。
  • 来源文件：
  • /inbox/tom/_candidates/2026-08-18-agent-rag-longcontext-candidates.json
  • [S2 enrich]