用Mamba模型生成长短结合的3D人体反应动作，速度快且真实多样。
ReactionMamba: Generating Short & Long Human Reaction Sequences
- 结合VAE与Mamba模型，高效编码解码动作序列。
- 在三组数据集上生成长序列动作，真实感与多样性优于现有方法。
- 推理速度显著提升，适合实时交互与复杂动作生成场景。
我们提出ReactionMamba，一种生成长时序3D人体反应动作的新框架。该框架采用运动变分自编码器（motion VAE）实现高效动作编码，并结合基于Mamba的状态空间模型进行时序一致的动作解码。这一设计使ReactionMamba能够生成从简单短动作到复杂长动作（如舞蹈、武术）的多样化反应序列。我们在NTU120-AS、Lindy Hop和InterX三个数据集上进行评估，结果表明，在动作真实性、多样性以及长序列生成能力方面，ReactionMamba相较InterFormer、ReMoS和Ready-to-React等方法表现优异，同时在推理速度上实现显著提升。
原文摘要 · Abstract (English)
We present ReactionMamba, a novel framework for generating long 3D human reaction motions. Reaction-Mamba integrates a motion VAE for efficient motion encoding with Mamba-based state-space models to decode temporally consistent reactions. This design enables ReactionMamba to generate both short sequences of simple motions and long sequences of complex motions, such as dance and martial arts. We evaluate ReactionMamba on three datasets--NTU120-AS, Lindy Hop, and InterX--and demonstrate competitive performance in terms of realism, diversity, and long-sequence generation compared to previous methods, including InterFormer, ReMoS, and Ready-to-React, while achieving substantial improvements in inference speed.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。