Immiscible Diffusion Policy: Preserving Multimodal Robot Actions through Label-Free Noise Assignment

📄 arXiv: 2610.09369v1 📥 PDF

作者: Xiao Zhang, Yuxin Chen, Zhixuan Liang, Guojian Zhan, Chenran Li, Chenfeng Xu, Masayoshi Tomizuka, Yiheng Li

分类: cs.RO, cs.LG

发布日期: 2026-10-07


💡 一句话要点

提出不可混溶扩散策略以解决机器人多模态动作保留问题

🎯 匹配领域: 支柱一:机器人控制 (Robot Control) 支柱二:RL算法与架构 (RL & Architecture) 支柱九:具身大模型 (Embodied Foundation Models)

关键词: 扩散策略 多模态学习 机器人规划 动作保留 无标签训练

📋 核心要点

  1. 现有的扩散策略在多模态动作分布恢复方面存在不足,常常导致模态崩溃为单一形式。
  2. 本文提出的不可混溶扩散策略通过无标签的动作噪声分配,旨在保留不同的噪声到动作路径。
  3. 实验结果显示,该方法在多模态任务中显著提高了非主导模态的比例,提升幅度达到6.0x-14.6x。

📝 摘要(中文)

在扩散策略首次提出时,期望其能够恢复多模态动作分布。然而,研究发现这一期望并不总是成立,因为扩散策略往往会崩溃为单一模态。分析表明,独立的动作噪声配对导致了扩散路径之间的混合和交叉,从而产生了平均去噪响应,抑制了模态特定行为。为了解决这一问题,本文提出了不可混溶扩散策略,这是一种无标签的训练时附加方法,通过动作噪声分配来保留相对独特的噪声到动作路径,而不修改策略架构或推理过程。实验结果表明,该方法在多项任务中显著提高了策略对动作模态的保留能力。

🔬 方法详解

问题定义:本文解决的问题是扩散策略在多模态动作分布恢复中的失败,现有方法在保证数据集模态平衡和批内对称性时,仍然容易崩溃为单一模态。

核心思路:提出不可混溶扩散策略,通过动作噪声分配来保持不同的噪声到动作路径,避免了独立动作噪声配对导致的混合和交叉,从而保留多模态特性。

技术框架:该方法在扩散策略的基础上进行改进,主要包括动作噪声的分配模块,确保在训练过程中不同模态的噪声路径相对独立。

关键创新:不可混溶扩散策略的核心创新在于其无标签的训练方式,能够在不改变策略架构或推理过程的情况下,增强多模态动作的保留能力。

关键设计:在参数设置上,采用了特定的噪声分配策略,损失函数设计上注重模态间的区分度,确保不同模态的动作能够得到有效保留。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果显示,采用不可混溶扩散策略后,非主导模态的比例在三项双模态任务中提升了6.0x-14.6x,并在四模态任务中恢复了在传统策略中完全缺失的模态。这表明该方法在多模态保留方面具有显著优势。

🎯 应用场景

该研究的潜在应用领域包括机器人规划、自动化操作和人机交互等场景。通过提高机器人在多模态任务中的表现,能够增强其在复杂环境中的适应能力,提升实际应用的价值和效率。未来,该方法可能推动更广泛的多模态学习研究和应用。

📄 摘要(原文)

When diffusion policies were first introduced, they were expected to recover multi-modal action distributions. However, we find this expectation does not always hold, as diffusion policies often collapse to a single modality even when we guarantee the balance of dataset modalities and exact within-batch symmetry. Our analysis indicates that independent action-noise pairing contributes to this failure by increasing mixing and crossing among diffusion paths, which can produce averaged denoising responses and suppress modality-specific behavior. This issue is especially severe in robot planning, where action spaces are dense and low-dimensional, significantly increasing such mixing and crossing. To alleviate this problem, we propose Immiscible Diffusion Policy, a label-free training-time add-on to diffusion policy that uses action-noise assignment to preserve relatively distinct noise-to-action routes without modifying the policy architecture or inference procedure. Across five simulated and two real-world humanoid manipulation tasks spanning state, RGB, and point-cloud observations, our method significantly improves the policy's preservation of action modalities while maintaining strong task performance. It increases the proportion of the non-dominant modality by 6.0x-14.6x across three two-modality tasks and recovers demonstrated modalities that are entirely absent from vanilla policy rollouts on both four-modality tasks. These results demonstrate that Immiscible Diffusion Policy provides a simple yet robust approach to preserving action multi-modality in general robot learning tasks.