InterControl: Zero-shot Human Interaction Generation by Controlling Every Joint
作者: Zhenzhi Wang, Jingbo Wang, Yixuan Li, Dahua Lin, Bo Dai
分类: cs.CV
发布日期: 2023-11-27 (更新: 2024-11-21)
备注: NeurIPS 2024 camera ready version. TL;DR: Generate human interactions with only single-person motion data in training via joint contact pairs from LLMs
🔗 代码/项目: GITHUB
💡 一句话要点
提出InterControl以解决多人人体交互生成问题
🎯 匹配领域: 支柱四:生成式动作 (Generative Motion) 支柱七:动作重定向 (Motion Retargeting) 支柱八:物理动画 (Physics-based Animation) 支柱九:具身大模型 (Embodied Foundation Models)
关键词: 运动合成 多人人体交互 零样本学习 扩散模型 逆向运动学 大型语言模型 动画生成
📋 核心要点
- 现有的运动生成方法主要针对单一角色,缺乏对多人人体交互的支持,限制了其应用场景。
- 本文提出InterControl,通过控制关节对之间的距离,实现任意数量角色的人体交互生成,具有灵活性和适应性。
- 实验结果显示,InterControl能够有效生成多角色交互,且与现有的物理模拟器兼容,展示了良好的应用潜力。
📝 摘要(中文)
文本条件下的运动合成在扩散模型的推动下取得了显著进展。然而,现有的运动扩散模型主要针对单一角色,忽视了多人人体交互的问题。本文提出了一种名为InterControl的方法,旨在以零样本方式合成任意数量角色的人体运动交互。该方法通过将人体交互建模为成对的关节,灵活地处理接触或保持特定距离的情况,超越了现有方法对训练数据的限制。实验结果表明,该框架能够生成多角色的交互,并与现成的物理基础角色模拟器兼容。
🔬 方法详解
问题定义:本文旨在解决现有运动生成模型在多人人体交互方面的不足,尤其是缺乏对任意数量角色的支持。现有方法通常需要在固定数量的角色数据集上进行训练,限制了其灵活性。
核心思路:InterControl的核心思路是将人体交互建模为关节对,通过控制关节之间的距离来实现自然的交互效果。这种设计允许模型在没有特定训练数据的情况下生成多角色的交互。
技术框架:该方法包括两个主要模块:运动控制器和逆向运动学引导模块。运动控制器负责生成关节运动,而逆向运动学模块确保生成的关节位置符合预期的距离和姿态。
关键创新:InterControl的创新在于其零样本生成能力,能够处理任意数量的角色交互,而不依赖于固定的训练数据集。这一特性使其在多角色交互生成中具有显著优势。
关键设计:在技术细节上,模型采用了特定的损失函数来优化关节对之间的距离,同时使用了现成的大型语言模型(LLM)来生成关节对的距离信息,确保生成的运动自然且符合物理规律。
🖼️ 关键图片
📊 实验亮点
实验结果表明,InterControl在生成多角色交互方面表现优异,相较于基线方法,生成的运动在自然性和准确性上有显著提升,能够有效处理复杂的交互场景。
🎯 应用场景
InterControl的潜在应用场景包括动画制作、游戏开发和虚拟现实等领域。其灵活的多角色交互生成能力能够大幅提升虚拟角色的表现力和互动性,具有重要的实际价值和未来影响。
📄 摘要(原文)
Text-conditioned motion synthesis has made remarkable progress with the emergence of diffusion models. However, the majority of these motion diffusion models are primarily designed for a single character and overlook multi-human interactions. In our approach, we strive to explore this problem by synthesizing human motion with interactions for a group of characters of any size in a zero-shot manner. The key aspect of our approach is the adaptation of human-wise interactions as pairs of human joints that can be either in contact or separated by a desired distance. In contrast to existing methods that necessitate training motion generation models on multi-human motion datasets with a fixed number of characters, our approach inherently possesses the flexibility to model human interactions involving an arbitrary number of individuals, thereby transcending the limitations imposed by the training data. We introduce a novel controllable motion generation method, InterControl, to encourage the synthesized motions maintaining the desired distance between joint pairs. It consists of a motion controller and an inverse kinematics guidance module that realistically and accurately aligns the joints of synthesized characters to the desired location. Furthermore, we demonstrate that the distance between joint pairs for human-wise interactions can be generated using an off-the-shelf Large Language Model (LLM). Experimental results highlight the capability of our framework to generate interactions with multiple human characters and its potential to work with off-the-shelf physics-based character simulators. Code is available at https://github.com/zhenzhiwang/intercontrol