arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2403.03181cs.LGcs.AIcs.RO
收藏

基于潜在动作的行为生成

Behavior Generation with Latent Actions

Seoul National University(首尔大学);New York University(纽约大学);Artificial Intelligence Institute of SNU(首尔大学人工智能研究院)

Seungjae Lee, Yibin Wang, Haritheja Etukuru, H. Jin Kim, Nur Muhammad Mahi Shafiullah, Lerrel Pinto

AI总结:

本文提出VQ-BeT模型,通过分层向量量化对连续动作进行分词,解决BeT在高维动作空间和长序列建模上的不足,在七个环境中超越现有方法并将推理速度提升5倍。

摘要:

从带标签数据集中对复杂行为进行生成建模一直是决策领域长期存在的难题。与语言或图像生成不同,决策需要建模动作——即分布呈多模态的连续值向量,这些动作可能来自未经筛选的数据源,且生成误差会在序列预测中不断累积。近期一类名为行为Transformer(BeT)的模型通过使用k-means聚类对动作进行离散化来捕捉不同模式,从而解决了这一问题。然而,k-means难以扩展到高维动作空间或长序列,并且缺乏梯度信息,因此BeT在建模长程动作方面表现不佳。在本工作中,我们提出了向量量化行为Transformer(VQ-BeT),这是一种用于行为生成的通用模型,能够处理多模态动作预测、条件生成和部分观测。VQ-BeT通过使用分层向量量化模块对连续动作进行分词来增强BeT。在包括仿真操作、自动驾驶和机器人技术在内的七个环境中,VQ-BeT相较于BeT和扩散策略等最先进模型均有提升。重要的是,我们证明了VQ-BeT在捕捉行为模式方面具有更强的能力,同时推理速度比扩散策略加快了5倍。视频和代码可在https://sjlee.cc/vq-bet获取。

英文摘要:

Generative modeling of complex behaviors from labeled datasets has been a longstanding problem in decision making. Unlike language or image generation, decision making requires modeling actions - continuous-valued vectors that are multimodal in their distribution, potentially drawn from uncurated sources, where generation errors can compound in sequential prediction. A recent class of models called Behavior Transformers (BeT) addresses this by discretizing actions using k-means clustering to capture different modes. However, k-means struggles to scale for high-dimensional action spaces or long sequences, and lacks gradient information, and thus BeT suffers in modeling long-range actions. In this work, we present Vector-Quantized Behavior Transformer (VQ-BeT), a versatile model for behavior generation that handles multimodal action prediction, conditional generation, and partial observations. VQ-BeT augments BeT by tokenizing continuous actions with a hierarchical vector quantization module. Across seven environments including simulated manipulation, autonomous driving, and robotics, VQ-BeT improves on state-of-the-art models such as BeT and Diffusion Policies. Importantly, we demonstrate VQ-BeT's improved ability to capture behavior modes while accelerating inference speed 5x over Diffusion Policies. Videos and code can be found https://sjlee.cc/vq-bet

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。

↑