arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2510.00726cs.ROcs.AIcs.LG
收藏

CroSTAta: 用于机器人操作的跨状态转换注意力变换器

CroSTAta: Cross-State Transition Attention Transformer for Robotic Manipulation

Istituto Italiano di Tecnologia(意大利技术研究院)

Giovanni Minelli, Giulio Turrisi, Victor Barasuol, Claudio Semini

AI总结:

CroSTAta通过引入状态转换注意力机制提升机器人操作策略的鲁棒性和时间推理能力。

摘要:

通过监督学习从示教中学习机器人操作策略仍然具有挑战性,当策略遇到训练过程中未明确涵盖的执行变化时。虽然通过注意力机制引入历史上下文可以提高鲁棒性,但标准方法在序列中处理所有过去状态,而没有显式建模演示可能包含的时序结构,如失败和恢复模式。我们提出了一种跨状态转换注意力变换器,采用一种新颖的状态转换注意力(STA)机制,根据学习的状态演变模式调节标准注意力权重,使策略能够更好地根据执行历史调整行为。我们的方法结合了这种结构化的注意力与训练期间的时间掩码,其中在最近的时间步中随机移除视觉信息,以鼓励从历史上下文中进行时间推理。在模拟评估中,STA在精度关键任务上比标准注意力方法和时间建模方法如TCN和LSTM网络表现更优,实现了超过2倍的改进。源代码和数据可通过https://github.com/iit-DLSLab/croSTAta访问。

英文摘要:

Learning robotic manipulation policies through supervised learning from demonstrations remains challenging when policies encounter execution variations not explicitly covered during training. While incorporating historical context through attention mechanisms can improve robustness, standard approaches process all past states in a sequence without explicitly modeling the temporal structure that demonstrations may include, such as failure and recovery patterns. We propose a Cross-State Transition Attention Transformer that employs a novel State Transition Attention (STA) mechanism to modulate standard attention weights based on learned state evolution patterns, enabling policies to better adapt their behavior based on execution history. Our approach combines this structured attention with temporal masking during training, where visual information is randomly removed from recent timesteps to encourage temporal reasoning from historical context. Evaluation in simulation shows that STA consistently outperforms standard attention approach and temporal modeling methods like TCN and LSTM networks, achieving more than 2x improvement over cross-attention on precision-critical tasks. The source code and data can be accessed at https://github.com/iit-DLSLab/croSTAta

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。

↑