无需训练，通过检索增强实现更通用的人体运动生成
RMD: A Simple Baseline for More General Human Motion Generation via Training-free Retrieval-Augmented Motion Diffuse
- 用外部检索库替代训练，灵活替换数据源
- 利用大模型拆分重组动作，支持身体部位复用
- 基于预训练扩散模型提升生成质量，适合处理罕见动作
尽管动作生成已取得显著进展，但其实际应用仍受限于数据集的多样性与规模，难以应对分布外场景。为此，我们提出一种简单有效的基准方法RMD，通过检索增强技术提升动作生成的泛化能力。与以往检索方法不同，RMD无需额外训练，具备三大优势：(1) 外部检索数据库可灵活替换；(2) 动作库中的身体部位可重复使用，由大语言模型辅助拆分与重组；(3) 利用预训练动作扩散模型作为先验，提升检索与直接组合所得动作的质量。无需训练，RMD在分布外数据上表现卓越，达到当前最优水平。
原文摘要 · Abstract (English)
While motion generation has made substantial progress, its practical application remains constrained by dataset diversity and scale, limiting its ability to handle out-of-distribution scenarios. To address this, we propose a simple and effective baseline, RMD, which enhances the generalization of motion generation through retrieval-augmented techniques. Unlike previous retrieval-based methods, RMD requires no additional training and offers three key advantages: (1) the external retrieval database can be flexibly replaced; (2) body parts from the motion database can be reused, with an LLM facilitating splitting and recombination; and (3) a pre-trained motion diffusion model serves as a prior to improve the quality of motions obtained through retrieval and direct combination. Without any training, RMD achieves state-of-the-art performance, with notable advantages on out-of-distribution data.
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。