InterAgent：基于交互图扩散的物理多智能体指令执行
InterAgent: Physics-based Multi-agent Command Execution via Diffusion on Interaction Graphs
提出交互图外感知表示与稀疏边注意力机制，可迁移到多智能体视觉感知与交互建模任务中，用于显式编码关节间空间依赖并动态剪枝冗余连接。
原文依据 we propose InterAgent, the first end-to-end framework for text-driven physics-based multi-agent humanoid control
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论人形智能体有望模拟人类社会行为中固有的复杂协调能力。然而，现有方法大多局限于单智能体场景，忽视了多智能体交互所必需的物理合理交互。为弥补这一差距，我们提出 InterAgent，这是首个用于文本驱动的物理多智能体人形控制的端到端框架。其核心是引入一种配备多流模块的自回归扩散 Transformer，该模块将本体感知、外感知和动作解耦，以减轻跨模态干扰，同时实现协同协调。我们进一步提出一种新颖的交互图外感知表示，显式捕获细粒度的关节间空间依赖关系，以促进网络学习。此外，我们在其中设计了一种基于稀疏边的注意力机制，动态剪枝冗余连接并强调关键的智能体间空间关系，从而增强交互建模的鲁棒性。大量实验表明，InterAgent 持续优于多个强基线，达到最优性能。它能够仅从文本提示生成连贯、物理合理且语义忠实的多智能体行为。项目页面：https://binlee26.github.io/InterAgent-Page
英文原文摘要 1431 字符
Humanoid agents are expected to emulate the complex coordination inherent in human social behaviors. However, existing methods are largely confined to single-agent scenarios, overlooking the physically plausible interplay essential for multi-agent interactions. To bridge this gap, we propose InterAgent, the first end-to-end framework for text-driven physics-based multi-agent humanoid control. At its core, we introduce an autoregressive diffusion transformer equipped with multi-stream blocks, which decouples proprioception, exteroception, and action to mitigate cross-modal interference while enabling synergistic coordination. We further propose a novel interaction graph exteroception representation that explicitly captures fine-grained joint-to-joint spatial dependencies to facilitate network learning. Additionally, within it we devise a sparse edge-based attention mechanism that dynamically prunes redundant connections and emphasizes critical inter-agent spatial relations, thereby enhancing the robustness of interaction modeling. Extensive experiments demonstrate that InterAgent consistently outperforms multiple strong baselines, achieving state-of-the-art performance. It enables producing coherent, physically plausible, and semantically faithful multi-agent behaviors from only text prompts. Project page: \tt \small \href https://binlee26.github.io/InterAgent-Page https://binlee26.github.io/InterAgent-Page .