arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2305.16612cs.RO
收藏

基于时空Transformer的强化学习用于机器人人群导航

Spatio-Temporal Transformer-Based Reinforcement Learning for Robot Crowd Navigation

Haodong He, Hao Fu, Qiang Wang, Shuai Zhou, Wei Liu

AI总结:

针对社交机器人导航中时空特征利用不足和策略保守的问题,提出基于时空Transformer的强化学习算法,通过门控嵌入和对齐特征,结合自适应熵,降低保守性并提升导航性能。

摘要:

社交机器人导航是一个开放且具有挑战性的问题。现有工作通常使用独立模块分别捕获空间和时间特征,但这种方法在提高时空特征利用率以及降低导航策略保守性方面带来了额外困难。为此,我们提出一种基于时空Transformer的策略优化算法,以增强时空特征的利用,从而促进人机交互的捕捉。具体而言,本文引入一种门控嵌入机制,通过在特征层面整合两种模态,有效对齐空间和时间表示。随后,利用Transformer对时空语义信息进行编码,以期找到最优导航策略。最后,将时空Transformer与自适应策略熵相结合,显著降低了导航策略的保守性。实验结果表明了所提框架的有效性,我们的方法展现出优越的性能。

英文摘要:

The social robot navigation is an open and challenging problem. In existing work, separate modules are used to capture spatial and temporal features, respectively. However, such methods lead to extra difficulties in improving the utilization of spatio-temporal features and reducing the conservative nature of navigation policy. In light of this, we present a spatio-temporal transformer-based policy optimization algorithm to enhance the utilization of spatio-temporal features, thereby facilitating the capture of human-robot interactions. Specifically, this paper introduces a gated embedding mechanism that effectively aligns the spatial and temporal representations by integrating both modalities at the feature level. Then Transformer is leveraged to encode the spatio-temporal semantic information, with hope of finding the optimal navigation policy. Finally, a combination of spatio-temporal Transformer and self-adjusting policy entropy significantly reduces the conservatism of navigation policies. Experimental results demonstrate the effectiveness of the proposed framework, where our method shows superior performance.

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。

↑