让物体响应人体动作，生成更真实的交互运动
HOI-Dyn: Learning Interaction Dynamics for Human-Object Motion Diffusion
- 把交互看作人驱动物的响应系统，用轻量Transformer建模动态反应
- 引入残差动力学损失，提升生成结果的物理合理性与因果一致性
- 适合关注动作生成真实性的研究人员，尤其在虚拟角色动画领域
生成真实的3D人-物交互（HOI）仍具挑战，因难以建模精细的交互动态。现有方法将人与物运动独立处理，导致物理上不成立且因果不一致的行为。本文提出HOI-Dyn，将HOI生成建模为驱动-响应系统，其中人体动作为驱动力，物体动作为响应。核心是基于轻量Transformer的交互动力学模型，显式预测物体对人类动作的反应。为进一步增强一致性，引入基于残差的动力学损失，缓解动力学预测误差的影响，防止误导优化信号。该动力学模型仅用于训练阶段，保障推理效率。通过大量定性和定量实验，验证了本方法不仅能提升HOI生成质量，还建立了一种可行的交互质量评估指标。
原文摘要 · Abstract (English)
Generating realistic 3D human-object interactions (HOIs) remains a challenging task due to the difficulty of modeling detailed interaction dynamics. Existing methods treat human and object motions independently, resulting in physically implausible and causally inconsistent behaviors. In this work, we present HOI-Dyn, a novel framework that formulates HOI generation as a driver-responder system, where human actions drive object responses. At the core of our method is a lightweight transformer-based interaction dynamics model that explicitly predicts how objects should react to human motion. To further enforce consistency, we introduce a residual-based dynamics loss that mitigates the impact of dynamics prediction errors and prevents misleading optimization signals. The dynamics model is used only during training, preserving inference efficiency. Through extensive qualitative and quantitative experiments, we demonstrate that our approach not only enhances the quality of HOI generation but also establishes a feasible metric for evaluating the quality of generated interactions.
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。