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~~

高校专区

University of Edinburgh(爱丁堡大学)

2026-03-02 至 2026-03-02 共收录 4
2511.10762 2026-03-02 cs.RO cs.CV

Attentive Feature Aggregation or: How Policies Learn to Stop Worrying about Robustness and Attend to Task-Relevant Visual Cues

关注特征聚合或:政策如何学会停止担心鲁棒性并关注任务相关的视觉线索

Nikolaos Tsagkas, Andreas Sochopoulos, Duolikun Danier, Sethu Vijayakumar, Alexandros Kouris, Oisin Mac Aodha, Chris Xiaoxuan Lu

机构 * University of Edinburgh(爱丁堡大学) UCL(伦敦大学学院) Samsung AI Center - Cambridge, UK(三星AI研究中心-剑桥,英国)

AI总结 本文提出AFA方法,通过注意力机制提升视觉-运动策略在扰动环境下的鲁棒性和泛化能力。

Comments This paper stems from a split of our earlier work "When Pre-trained Visual Representations Fall Short: Limitations in Visuo-Motor Robot Learning." While "The Temporal Trap" replaces the original and focuses on temporal entanglement, this companion study examines policy robustness and task-relevant visual cue selection. arXiv admin note: text overlap with arXiv:2502.03270

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.05535 2026-03-02 cs.RO cs.CL cs.HC cs.LG cs.MA

Mixed-Initiative Dialog for Human-Robot Collaborative Manipulation

混合发起对话用于人机协作操作

Albert Yu, Chengshu Li, Luca Macesanu, Arnav Balaji, Ruchira Ray, Raymond Mooney, Roberto Martín-Martín

机构 * UT Austin(得克萨斯大学) OpenAI NYU(纽约大学) University of Edinburgh(爱丁堡大学)

AI总结 MICoBot通过混合发起对话范式提升人机协作任务的成功率和用户体验。

Comments Project website at https://robin-lab.cs.utexas.edu/MicoBot/

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.16607 2026-03-02 cs.LG cs.AI

Asymptotically Stable Quaternion-valued Hopfield-structured Neural Network with Periodic Projection-based Supervised Learning Rules

渐近稳定的四元数值Hopfield结构神经网络及其基于周期投影的监督学习规则

Tianwei Wang, Xinhui Ma, Wei Pang

机构 * University of Edinburgh(爱丁堡大学) University of Hull(赫尔大学) Heriot-Watt University(赫瑞-沃德大学)

AI总结 本文提出了一种基于四元数的Hopfield结构神经网络,通过周期投影策略实现监督学习,具有高精度、快速收敛和强可靠性,适用于机器人控制等需要四元数参数化的场景。

Comments Preprint. Accepted by NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2507.02965 2026-03-02 cs.CV cs.AI

Concept-based Adversarial Attack: a Probabilistic Perspective

基于概念的对抗攻击:概率视角

Andi Zhang, Xuan Ding, Steven McDonagh, Samuel Kaski

机构 * University of Warwick(沃里克大学) University of Manchester(曼彻斯特大学) The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳)) University of Edinburgh(爱丁堡大学) University of Aalto(阿alto大学)

AI总结 本文提出一种基于概念的对抗攻击方法,通过概率视角生成多样化的对抗示例,保持原始概念以误导分类器,提升攻击效率。

详情

展开后加载摘要…

URL PDF HTML 收藏