arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.01812cs.ROcs.LG
收藏

等变扩散策略

Equivariant Diffusion Policy

Boston Dynamics AI Institute(波士顿动力人工智能研究所);Northeastern University(东北大学)

Dian Wang, Stephen Hart, David Surovik, Tarik Kelestemur, Haojie Huang, Haibo Zhao, Mark Yeatman, Jiuguang Wang, Robin Walters, Robert Platt

AI总结:

提出 Equivariant Diffusion Policy,利用领域对称性提升去噪函数的样本效率与泛化能力,在 6-DoF 控制任务中显著优于基线 Diffusion Policy。

摘要:

近期工作表明,扩散模型是学习行为克隆中示范数据产生的多模态分布的有效方法。然而,该方法的缺点是需要学习一个去噪函数,这比学习显式策略要复杂得多。在本研究中,我们提出 Equivariant Diffusion Policy,这是一种新颖的扩散策略学习方法,利用领域对称性在去噪函数中获得更好的样本效率和泛化能力。我们从理论上分析了完整 6-DoF 控制的 $\mathrm{SO}(2)$ 对称性,并刻画了扩散模型何时具有 $\mathrm{SO}(2)$-等变性。我们进一步在 MimicGen 中的 12 项模拟任务上对方法进行了实证评估,结果表明其成功率平均比基线 Diffusion Policy 高 21.9%。我们还在真实系统上评估了该方法,表明可以用相对较少的训练样本学习到有效策略,而基线 Diffusion Policy 则无法做到。

英文摘要:

Recent work has shown diffusion models are an effective approach to learning the multimodal distributions arising from demonstration data in behavior cloning. However, a drawback of this approach is the need to learn a denoising function, which is significantly more complex than learning an explicit policy. In this work, we propose Equivariant Diffusion Policy, a novel diffusion policy learning method that leverages domain symmetries to obtain better sample efficiency and generalization in the denoising function. We theoretically analyze the $\mathrm{SO}(2)$ symmetry of full 6-DoF control and characterize when a diffusion model is $\mathrm{SO}(2)$-equivariant. We furthermore evaluate the method empirically on a set of 12 simulation tasks in MimicGen, and show that it obtains a success rate that is, on average, 21.9% higher than the baseline Diffusion Policy. We also evaluate the method on a real-world system to show that effective policies can be learned with relatively few training samples, whereas the baseline Diffusion Policy cannot.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑