Mitigating Error Accumulation in Co-Speech Motion Generation via Global Rotation Diffusion and Multi-Level Constraints
通过全局旋转扩散和多级约束缓解语音同步运动生成中的误差累积
专题命中 扩散模型 :diffusion(title,abstract);分类 cs.CV
AI总结 本文提出GlobalDiff,通过全局旋转扩散和多级约束缓解语音同步运动生成中的误差累积,提升生成运动的平滑度和准确性。
Comments AAAI 2026. Project page: this https URL (https://xiangyuezhang.com/GlobalDiff/;) code and pretrained models: this https URL (https://github.com/Xiangyue-Zhang/GlobalDiff)