IROS - arXivDaily 期刊&会议

2606.20491 2026-06-19 cs.RO cs.CV 新提交

Fast Human Attention Prediction for Fixation-guided Active Perception in Autonomous Navigation

用于自主导航中注视引导主动感知的快速人类注意力预测

Fatma Youssef Mohammed, Grzegorz Malczyk, Kostas Alexis

机构 * Norwegian University of Science and Technology (NTNU)（挪威科技大学）

AI总结提出GazeLNN，一种基于液态神经网络和MobileNetV3的轻量级扫描路径预测模型，在MIT低分辨率数据集上达到最优性能，计算成本降低99.40%，推理速度提升6倍，并集成到强化学习训练的主动相机-机器人控制策略中，实现自主导航中的注视引导感知。

Comments Accepted to the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)

2606.20426 2026-06-19 cs.RO 新提交

TaCauchy: An Extensible FEM Framework for Vision-Based Tactile Simulation

TaCauchy：面向视觉触觉仿真的可扩展有限元框架

Hengfei Zhao, Yifan Xie, Junhao Gong, Yue Sun, Kai Zhu, Weihua He, Shoujie Li, Haohuan Fu, Wenbo Ding

机构 * Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院）； Huawei Inc.（华为技术有限公司）

AI总结提出TaCauchy框架，基于UIPC求解器在Isaac Sim中集成有限元法，直接计算柯西应力张量并投影为接触力，实现高保真触觉仿真，支持多种传感器，物理验证SSIM>0.93。

Comments Accepted to IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2026

2606.20424 2026-06-19 cs.RO 新提交

LIT-GS: LiDAR-Inertial-Thermal Gaussian Splatting for Illumination-Robust Mapping

LIT-GS: 面向光照鲁棒建图的激光雷达-惯性-热高斯泼溅

Shikuan Shi, Chunran Zheng, Jiaming Xu, Tianyong Ye, Tao Yu, Yukang Cui

机构 * College of Mechatronics and Control Engineering, Shenzhen University（深圳大学机电与控制工程学院）； Department of Mechanical Engineering, The University of Hong Kong（香港大学机械工程系）

AI总结提出LIT-GS框架，利用激光雷达平面几何约束联合优化位姿与高斯，解决光照变化和纹理缺失场景下RGB依赖的脆弱性问题，提升几何精度与渲染质量。

Comments Accepted to IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)

2606.20056 2026-06-19 cs.RO 新提交

VFILC: Accurate Frequency Extrapolations in Imitation Learning via Sampling Frequency ILC

VFILC: 通过采样频率迭代学习控制实现模仿学习中的精确频率外推

Nozomu Masuya, Toshiaki Tsuji, Sho Sakaino

机构 * Grad. School of Science ； Technology University of Tsukuba Tsukuba, Japan ； Engineering Saitama University Saitama, Japan ； Information Engineering University of Tsukuba Tsukuba, Japan

AI总结提出VFILC方法，结合可变频率模仿学习与前馈-反馈迭代学习控制，在三种任务中实现精确的速度外推，频率误差降低最高81%。

Comments 8 pages, 17 figures. Accepted at IROS 2026

2606.19632 2026-06-19 cs.RO cs.AI cs.LG cs.LO cs.MA 新提交

Formal Verification of Learned Multi-Agent Communication Policies via Decision Tree Distillation

通过决策树蒸馏对学习到的多智能体通信策略进行形式化验证

Ahmad Farooq, Kamran Iqbal

机构 * University of Arkansas at Little Rock（阿肯色大学小石城分校）

AI总结提出通过决策树蒸馏将多智能体强化学习策略转化为可解释模型，并利用PRISM进行形式化验证，确保安全属性转移至原始网络，在无人机编队任务中实现88.9%属性满足率。

Comments 9 pages, 3 figures, 7 tables. Accepted at the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026), Pittsburgh, Pennsylvania, USA, September 27-October 1, 2026