arXivDaily arXiv每日学术速递 周一至周五更新

期刊&会议

International Conference on Intelligent Robots and Systems · 会议 · Robotics

共收录 116
2608.10383 2026-08-14 cs.RO 版本更新

Real-World Cooperative Bimanual Dexterous Grasp of Large Objects from Single-View Observations

基于单视图观测的真实场景中大型物体的双臂协同灵巧抓取

Ziming Li, Mingxuan Wu, Jiaqi Zhang, Hongfei Li, Yan Gan, Deqiang Ouyang, Ning Wang

机构 * The University of Auckland(奥克兰大学) Chongqing University(重庆大学) Southwest University(西南大学)

AI总结 本文针对机器人双臂抓取大型物体的挑战,提出含多模态数据集、DDPM模块及执行策略的真实场景双臂抓取框架,实验表明其对未见物体抓取成功率高。

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2608.00237 2026-08-13 cs.CV cs.RO 版本更新

Latent-Centroid Steering: Single-Pass Classifier-Free Guidance for Command-Aligned Autonomous Driving

隐式质心引导:用于指令对齐自动驾驶的单次无分类器引导

Meibo Hu, Jiamian Wang, Pichao Wang, Zhiqiang Tao

机构 * Rochester Institute of Technology(罗切斯特理工学院) NVIDIA(英伟达公司)

AI总结 针对视觉语言自动驾驶模型的指令跟随差距,提出单次引导机制LCS,降低推理延迟约50%,在Bench2Drive和nuScenes基准上提升指令遵循与驾驶性能。

Comments IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.24060 2026-08-13 cs.RO 版本更新

RoboHarness: A Memory-Augmented Policy Harness for Vision-Language-Action Model Robustness via In-Context Adaptation

SOMA:通过上下文适应提升视觉-语言-动作模型鲁棒性的战略编排与内存增强系统

Zhuoran Li, Zhiyang Li, Kaijun Zhou, Jinyu Gu

AI总结 SOMA通过对比双记忆检索增强生成(RAG)、归因驱动大语言模型(LLM)编排器和可扩展模型上下文协议(MCP)干预,提升视觉-语言-动作模型在分布外任务中的鲁棒性,实验表明其在长周期任务链中提升了89.1%的绝对成功率。

Comments 8 pages, 10 figures, 4 tables. Accepted to the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026). Project page and source code: https://github.com/LZY-1021/RoboHarness

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.21165 2026-08-11 cs.RO cs.AI 版本更新

OmniV2X: A Generative Foundation Planner for Efficient End-to-End Cooperative Driving

OmniV2X:一种用于高效端到端协同驾驶的生成式基础规划器

Juntong Peng, Juanwu Lu, Yupeng Zhou, Can Cui, Yaobin Chen, Ziran Wang

机构 * Purdue University(普渡大学)

AI总结 提出OmniV2X生成式基础模型,通过端到端监督训练和交叉注意力注入,利用多模态多智能体上下文实现高效协同驾驶,在DAIR-V2X-Seq数据集上以少于10%微调数据和1%通信带宽达到最优性能。

Comments Accepted to IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.05868 2026-08-11 cs.RO 版本更新

AnyCamVLA: Zero-Shot Camera Adaptation for Viewpoint Robust Vision-Language-Action Models

AnyCamVLA: 零样本相机适应用于视角鲁棒的视觉-语言-动作模型

Hyeongjun Heo, Seungyeon Woo, Sang Min Kim, Junho Kim, Junho Lee, Yonghyeon Lee, Young Min Kim

机构 * Department of Electrical and Computer Engineering, Seoul National University(电子与计算机工程系,首尔国立大学) Department of Mechanical Engineering, Massachusetts Institute of Technology(机械工程系,麻省理工学院)

AI总结 AnyCamVLA通过零样本相机适应提升视觉-语言-动作模型在视角变化下的鲁棒性,适用于任何RGB策略。

Comments Accepted to IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2607.18540 2026-08-07 cs.CV cs.LG cs.RO 版本更新

Recti-Q: Feature-Space Rectification for Out-of-Distribution-Robust Quantized Perception in Edge Robotics

Recti-Q:用于边缘机器人中分布外鲁棒量化感知的特征空间校正

Hamidreza Yaghoubi Araghi, Parastoo Pilevar, Ming C. Lin

AI总结 研究针对边缘机器人中PTQ在分布变化下可靠性降低的问题,提出轻量级特征空间校正框架Recti-Q,它冻结量化主干,仅用源数据训练小型分类器头LoRA适配器,与架构无关,能恢复鲁棒性,节省内存且实现低带宽OTA弹性修补。

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.19229 2026-08-07 cs.RO cs.AI cs.CV cs.LG cs.SY eess.SY 版本更新

NavTrust: Benchmarking Trustworthiness for Embodied Navigation

NavTrust:体感导航的可信度基准测试

Huaide Jiang, Yash Chaudhary, Yuping Wang, Zehao Wang, Raghav Sharma, Manan Mehta, Yang Zhou, Lichao Sun, Zhiwen Fan, Zhengzhong Tu, Jiachen Li

机构 * Trustworthy Autonomous Systems Laboratory at the University of California, Riverside(加州大学河滨分校可信自主系统实验室) University of Michigan(密歇根大学) Workday University of Southern California(南加州大学) Texas A&M University(德克萨斯A&M大学) Lehigh University(莱斯大学)

AI总结 NavTrust通过系统性地对输入模态进行腐蚀,评估体感导航性能的鲁棒性,揭示了现有方法在真实环境中的性能缺陷,并提出改进策略。

Comments IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026); Project Website: https://navtrust.github.io

详情

展开后加载摘要…

URL PDF HTML 收藏
2607.18660 2026-08-05 cs.RO 版本更新

MVP-Tac: A Miniaturized Dual-Modal Vision and Photoelastic Tactile Sensor for Robot-Assisted Minimally Invasive Surgery

MVP-Tac:一种用于机器人辅助微创手术的小型化双模态视觉与光弹性触觉传感器

Md Rakibul Islam Prince, Jaeeun Kim, Yuhao Zhou, Mason Vrshek, Shivani Reddy Sama, Adyaa Khera, Sheeraz Athar, Zijie Xu, Jiabin Liu, Shaoting Lin, Wei Li, Yu She

机构 * Elmore Family School of Electrical and Computer Engineering, Purdue University(普渡大学埃尔莫尔电气与计算机工程学院) Edwardson School of Industrial Engineering, Purdue University(普渡大学爱德华森工业工程学院) School of Mechanical Engineering, Purdue University(普渡大学机械工程学院) Robotics Engineering Technology, Purdue University(普渡大学机器人工程技术学院) Department of Civil Engineering, Stony Brook University(石溪大学土木工程系) Department of Mechanical Engineering, Michigan State University(密歇根州立大学机械工程系)

AI总结 研究针对机器人辅助微创手术缺乏触觉反馈问题,介绍了MVP-Tac传感器,它采用反射光弹性成像,能在视觉触觉间切换,经力校准及肿瘤触诊等实验验证其有效性,为恢复RMIS中触诊并保持视觉反馈提供实用途径。

Comments 8 pages, 8 figures. To appear in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.14068 2026-08-05 cs.RO 版本更新

Stiffness Copilot: An Impedance Policy for Contact-Rich Teleoperation

刚性助手:一种用于接触丰富的远程操作的阻抗策略

Yeping Wang, Zhengtong Xu, Pornthep Preechayasomboon, Ben Abbatematteo, Amirhossein H. Memar, Nick Colonnese, Sonny Chan

AI总结 本文提出Stiffness Copilot,一种基于视觉的共享控制策略,通过在线调整机器人刚度实现安全高效的接触丰富任务远程操作。

Comments Accepted to IROS 2026. Project website: https://stiffness-copilot.github.io

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.03270 2026-08-04 cs.RO cs.AI 版本更新

Grounded Vision-Language Interpreter for Long-Horizon Bimanual Task and Motion Planning

用于长 horizon 双臂任务与运动规划的接地视觉语言解释器

Jeremy Siburian, Keisuke Shirai, Cristian C. Beltran-Hernandez, Masashi Hamaya, Michael Görner, Atsushi Hashimoto

机构 * OMRON SINIC X Corporation(OMRON SINIC X公司) The University of Tokyo(东京大学) University of Hamburg(汉堡大学)

AI总结 针对现有视觉语言机器人规划框架的黑箱缺陷与双臂任务探索不足问题,提出混合规划框架 ViLaIn-TAMP,经烹饪领域任务及实际双臂机器人系统验证,其性能优于基准方法。

Comments IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.14431 2026-08-04 cs.RO 版本更新

Local-Canonicalization Equivariant Graph Neural Networks for Sample-Efficient and Generalizable Swarm Robot Control

用于样本高效且可泛化的群体机器人控制的局部正则化等变图神经网络

Keqin Wang, Tao Zhong, David Chang, Christine Allen-Blanchette

机构 * Princeton University(普林斯顿大学)

AI总结 该研究针对群体控制MARL策略效率低、泛化差的问题,提出LEGO架构,结合正则化与角色感知图编码,搭配MAPPO算法,在多基准测试中提升性能,可跨团队规模迁移,Crazyflie实验中失效后仍可运行。

Comments Accepted at IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2409.07163 2026-08-04 cs.RO cs.CV 版本更新

Mamba Policy: Towards Efficient 3D Diffusion Policy with Hybrid Selective State Models

Mamba Policy:基于混合选择性状态模型的高效3D扩散策略

Jiahang Cao, Qiang Zhang, Jingkai Sun, Jiaxu Wang, Hao Cheng, Yulin Li, Jun Ma, Kun Wu, Zhiyuan Xu, Yecheng Shao, Wen Zhao, Gang Han, Yijie Guo, Renjing Xu

机构 * Microelectronics Thrust, The Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州)微电子领域) Division of Emerging Interdisciplinary Areas, The Hong Kong University of Science and Technology(香港科技大学新兴交叉领域 division) Beijing Innovation Center of Humanoid Robotics(北京人形机器人创新中心) Center for X-Mechanics, Zhejiang University(浙江大学X力学中心)

AI总结 本研究提出Mamba Policy,以XMamba Block融合Mamba与注意力机制,参数量减超80%,在Adroit等数据集上性能优异且计算资源需求低,长 horizon 场景鲁棒性更强。

Comments Accepted to IROS 2025. Project Page: https://sagecao1125.github.io/mamba_policy/

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.14584 2026-07-31 cs.RO 版本更新

A Robust Placeability Metric for Model-Free Unified Pick-and-Place Reasoning

一种针对无模型统一抓取放置推理的鲁棒放置性度量

Benno Wingender, Nils Dengler, Rohit Menon, Sicong Pan, Maren Bennewitz

机构 * Anonymous Authors(匿名作者) Benno Wingender(伯恩诺·温格德尔) Nils Dengler(尼尔·登格尔) Rohit Menon(罗希特·梅农) Sicong Pan(斯冰·潘) Maren Bennewitz(马伦·本内维茨)

AI总结 本文提出一种鲁棒的概率放置性度量,通过联合评估稳定性、抓取性和空隙,实现无模型的统一抓取放置推理,并在仿真和真实机器人实验中验证其有效性。

Comments IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.11970 2026-07-31 cs.RO 版本更新

ARCSnake V2: Mechanical Adaptations For An Amphibious Multi-Domain Screw-Propelled Snake-Like Robot

ARCSnake V2:适用于水陆两栖多领域的螺旋推进蛇形机器人的机械改进

Sara Wickenhiser, Lizzie Peiros, Calvin Joyce, Peter Gavrilov, Sujaan Mukherjee, Syler Sylvester, Junrong Zhou, Mandy Cheung, Jason Lim, Florian Richter, Michael C. Yip

机构 * Mechanical and Aerospace Engineering Department, University of California San Diego(加州大学圣地亚哥分校机械与航空航天工程系) Electrical and Computer Engineering Department, University of California, San Diego(加州大学圣地亚哥分校电气与计算机工程系)

AI总结 ARCSnake V2是ARCSnake V1的改进型两栖螺旋推进蛇形机器人,结合超冗余蛇形机器人高机动性与阿基米德螺旋推进的地形适应性,经实验验证具备水下作业能力,可作为多领域探索等任务的多功能平台。

Comments 8 pages, 5 figures, IROS

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.04999 2026-07-31 cs.RO cs.AI cs.LG 版本更新

CLAM: Continuous Latent Action Models for Robot Learning from Unlabeled Demonstrations

CLAM:基于未标记演示的连续潜在动作模型用于机器人学习

Anthony Liang, Pavel Czempin, Matthew M. Hong, Yutai Zhou, Jingzhen Wang, Erdem Biyik, Stephen Tu

机构 * Department of Computer Science, University of Southern California(南加州大学计算机科学系) Department of Electrical and Computer Engineering, University of Southern California(南加州大学电气与计算机工程系)

AI总结 CLAM通过连续潜在动作标签和联合训练动作解码器,有效解决复杂连续控制任务中未标记数据的学习问题,实现在DMControl和MetaWorld等基准及真实机器人上的性能提升。

Comments Latent Action Models, Self-supervised Pretraining, Learning from Videos

Journal ref IEEE/RSJ International Conference on Intelligent Robots and Systems 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2602.09472 2026-07-30 cs.RO cs.CV 版本更新

LLM-Grounded Dynamic Task Planning with Hierarchical Temporal Logic for Human-Aware Multi-Robot Handover

基于分层时序逻辑的LLM引导动态任务规划用于人感知多机器人协作

Shuyuan Hu, Tao Lin, Kai Ye, Tianwei Zhang

机构 * The Shenzhen Institute of Artificial Intelligence and Robotics for Society(深圳人工智能与机器人社会研究院) Harbin Institute of Technology(哈尔滨工业大学) The Chinese University of Hong Kong-Shenzhen(香港中文大学(深圳)) Tsinghua Shenzhen International Graduate School(清华大学深圳国际 Graduate School)

AI总结 本文提出基于分层时序逻辑的神经符号框架,实现动态任务规划以提升多机器人协作的效率和鲁棒性。

Comments Accepted by IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.26017 2026-07-28 cs.RO 版本更新

G2DP: Diffusion Planning with Spatio-Temporal Grid Guidance

G2DP: 基于时空网格引导的扩散规划

Hang Yu, Ye Jin, Alessandro Canevaro, Julian Schmidt, Julian Jordan, Peizheng Li, Marc Kaufeld, Silvan Lindner, Johannes Betz, Wilhelm Stork

机构 * Mercedes-Benz AG(梅赛德斯-奔驰集团) Karlsruhe Institute of Technology(卡尔斯鲁厄理工学院) TU Munich(慕尼黑工业大学) University of Tübingen(图宾根大学)

AI总结 针对自动驾驶扩散规划器随机性导致的安全与路线保持问题,提出G2DP,通过可微时空代价体积在去噪过程中注入密集梯度,实现无碰撞与路径最优的轨迹生成,在nuPlan等基准上取得最优性能。

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.25897 2026-07-28 cs.RO cs.LG cs.SY eess.SY 版本更新

Variational Neural Belief Parameterizations for Robust Dexterous Grasping under Multimodal Uncertainty

变分神经信念参数化用于多模态不确定性下的鲁棒灵巧抓取

Clinton Enwerem, Shreya Kalyanaraman, John S. Baras, Calin Belta

机构 * Department of Electrical & Computer Engineering and Institute for Systems Research, University of Maryland, College Park, MD, USA(电气与计算机工程系和系统研究所,马里兰大学,College Park, MD, USA) Maryland Applied Graduate Engineering, A. James Clark School of Engineering, University of Maryland, College Park, MD, USA(马里兰应用研究生工程学院,A. James Clark工程学院,马里兰大学,College Park, MD, USA)

AI总结 本文提出变分推理方法,通过可微高斯混合模型表示信念,利用Gumbel-Softmax和位置-尺度重参数化实现平滑采样,提升抓取鲁棒性并减少规划时间。

Comments 11 pages, 10 figures. Accepted for publication at IROS 2026. Code, simulation assets, and dataset at https://github.com/coenwerem/vnb-grasp

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.08521 2026-07-28 cs.CV cs.RO eess.IV 版本更新

OccTrack360: 4D Panoptic Occupancy Tracking from Surround-View Fisheye Cameras

OccTrack360: 从环视鱼眼相机实现4D全景占用跟踪

Yongzhi Lin, Kai Luo, Yuanfan Zheng, Hao Shi, Mengfei Duan, Yang Liu, Kailun Yang

机构 * School of Artificial Intelligence and Robotics, Hunan University(人工智能与机器人学院,湖南大学) State Key Laboratory of Extreme Photonics and Instrumentation, Zhejiang University(极端光子学与仪器国家重点实验室,浙江大学) National Engineering Research Center of Robot Visual Perception and Control Technology, Hunan University(机器人视觉感知与控制技术国家工程研究中心,湖南大学)

AI总结 OccTrack360提出新的4D全景占用跟踪基准及FoSOcc框架,解决鱼眼成像中的球面投影和体素定位问题,提升占用跟踪性能。

Comments Accepted to IEEE/RSJ IROS 2026. The benchmark and source code will be made publicly available at https://github.com/YouthZest-Lin/OccTrack360

详情

展开后加载摘要…

URL PDF HTML 收藏
2607.20679 2026-07-27 cs.RO 版本更新

Towards Capability-Aware Traversability Navigation for Unstructured Environments

面向非结构化环境的能力感知可通行性导航

Gianluca Capezzuto, Felipe Tommaselli, Matheus P. Angarola, Ricardo V. Godoy, Marcelo Becker

机构 * University of São Paulo(圣保罗大学)

AI总结 研究非结构化环境中可通行性估计,提出能力感知可通行性(CAT)框架,将物理限制嵌入空间特征空间,通过交互式标注和SPADE块改进可通行性预测,在多数据集上领先,实现实例感知避障。

Comments 8 pages, 7 figures. Accepted to the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026). Project page: https://capability-aware-traversability.github.io/

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.18336 2026-07-24 cs.RO cs.CV 版本更新

Enhancing Glass Surface Reconstruction via Depth Prior for Robot Navigation

通过深度先验增强玻璃表面重建以提升机器人导航

Jiamin Zheng, Jingwen Yu, Guangcheng Chen, Hong Zhang

机构 * Shenzhen Key Laboratory of Robotics and Computer Vision(机器人与计算机视觉深圳重点实验室) Southern University of Science and Technology(南方科技大学) CKS Robotics Institute(CKS机器人研究院)

AI总结 本文提出一种无需训练的框架,利用深度基础模型作为结构先验,结合鲁棒的局部RANSAC对齐方法,融合原始传感器深度数据,从而避免错误的玻璃测量污染,恢复准确的度量尺度,并引入新的RGB-D数据集用于玻璃区域的几何真实值。

Comments 9 pages, 8 figures, Accepted by IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.04038 2026-07-24 cs.RO 版本更新

Force-Aware Residual DAgger via Trajectory Editing for Precision Insertion with Impedance Control

具备力感知的残差数据集聚合轨迹编辑用于阻抗控制的精确插入

Yiou Huang, Ning Ma, Weichu Zhao, Zinuo Liu, Jun Sun, Qiufeng Wang, Yaran Chen

机构 * Xi’an Jiaotong-Liverpool University(西交利物浦大学)

AI总结 本文提出TER-DAgger框架,通过优化轨迹编辑学习残差策略,减少协变量偏移,结合力感知的失败预判机制和阻抗控制框架,提升精确插入任务的成功率。

Comments 8 pages, 3 figures. Accepted to the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2607.17332 2026-07-23 cs.RO 版本更新

Multi-Resolution Voxelized Map-Based Stereo Visual-Inertial Odometry

基于多分辨率体素化地图的立体视觉惯性里程计

Shuyi Pan, Hangtian Wang, Zhaoxing Zhang, Chengliang Zhang, Zikang Yuan, Xin Yang

机构 * Huazhong University of Science and Technology(华中科技大学) AI Chip Center for Emerging Smart Systems (AC-CESS)(新兴智能系统人工智能芯片中心)

AI总结 研究视觉惯性里程计中姿态估计问题,提出多分辨率先验地图构建方法及基于地图的VIO系统,通过体素化地图、锥形索引策略和DDA算法,减少数据传输量与计算负载,实验证明该系统能在少数据传输下实现准确姿态估计。

Comments 9 pages, 4 figures, 6 tables. Accepted to appear in the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.19031 2026-07-23 cs.RO 版本更新

Congestion-Aware Robot Tour Planning in Crowded Environments

拥挤环境中的拥塞感知机器人巡视规划

Stefano Bernagozzi, Charlie Street, Masoumeh Mansouri, Lorenzo Natale

机构 * Istituto Italiano di Tecnologia(意大利理工学院) Università di Genova(热那亚大学) University of Birmingham(伯明翰大学)

AI总结 提出一种基于概率的巡视规划器,通过学习人流预测模型并在线构建马尔可夫决策过程,在拥挤环境中高效规划机器人路径,减少拥塞影响。

Comments Accepted to IEEE IROS 2026

Journal ref IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2605.31119 2026-07-23 cs.RO cs.LG 版本更新

Don't Fool Me Twice: Adapting to Adversity in the Wild with Experience-Driven Reasoning

不要愚弄我两次:通过经验驱动推理在野外适应逆境

Navin Sriram Ravie, Andrew Jong, Krrish Jain, John Liu, Omar Alama, Bijo Sebastian, Sebastian Scherer

机构 * Department of Engineering Design, Indian Institute of Technology, Madras(印度理工学院工程设计系,马德拉斯) Robotics Institute, Carnegie Mellon University(卡内基梅隆大学机器人研究所)

AI总结 提出一种持续学习框架,使移动机器人能够在线从干扰中学习,通过语义将异常行为归因于原因,从而更好地预测和规划未来。

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.28029 2026-07-23 cs.CV cs.RO 版本更新

Effort-Based Criticality Metrics for Evaluating 3D Perception Errors in Autonomous Driving

基于努力的临界度量指标用于评估自动驾驶中的3D感知误差

Sharang Kaul, Simon Bultmann, Mario Berk, Abhinav Valada

机构 * CARIAD SE University of Freiburg(弗莱堡大学)

AI总结 本文提出基于努力的临界度量指标,通过FSR、MDR和LEA量化自动驾驶中3D感知误差的严重性,揭示非关键性误差占比高,验证新指标能捕捉安全关键信息。

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.06210 2026-07-23 cs.CV cs.RO 版本更新

VG3S: Visual Geometry Grounded Gaussian Splatting for Semantic Occupancy Prediction

VG3S:基于视觉几何的高斯散射用于语义占用预测

Xiaoyang Yan, Muleilan Pei, Shaojie Shen

机构 * Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology(电子与计算机工程系,香港科学与技术大学)

AI总结 VG3S 通过引入视觉基础模型的几何 grounding 能力,提升语义占用预测的准确性与泛化能力。

Comments Accepted by IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2602.04401 2026-07-23 cs.RO cs.CV 版本更新

Quantile Transfer for Reliable Operating Point Selection in Visual Place Recognition

视觉地点识别中可靠操作点选择的分位数迁移

Dhyey Manish Rajani, Michael Milford, Tobias Fischer

机构 * QUT Centre for Robotics(昆士兰理工大学机器人中心) School of Electrical Engineering and Robotics(电气工程与机器人学院) Queensland University of Technology(昆士兰理工大学)

AI总结 提出一种通过分位数归一化迁移阈值的方法,自动选择视觉地点识别系统的操作点,在100%精度下最大化召回率,无需手动调参。

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.18348 2026-07-23 cs.RO cs.AI cs.LG 版本更新

PGTT: Phase-Guided Terrain Traversal for Perceptive Legged Locomotion

PGTT:用于感知式有腿运动的相位引导地形遍历

Alexandros Ntagkas, Chairi Kiourt, Konstantinos Chatzilygeroudis

机构 * Laboratory of Automation and Robotics (LAR) in the Department of Electrical & Computer Engineering, University of Patras(自动化与机器人实验室(LAR)(电气与计算机工程系,帕特拉斯大学)) Archimedes/Athena RC, Greece(阿基米德/雅典娜RC,希腊) Athena - Research and Innovation Center in Information, Communication and Knowledge Technologies, Xanthi, Greece(雅典娜信息、通信和知识技术研究中心(Xanthi,希腊)) Computational Intelligence Laboratory (CILab), Department of Mathematics, University of Patras(计算智能实验室(CILab)(数学系,帕特拉斯大学))

AI总结 研究针对有腿机器人感知强化学习控制器的不足,提出相位引导地形遍历(PGTT)方法,通过奖励塑造强化步态结构,编码腿相位等,在特定训练下,该方法在多种场景成功率高且速度跟踪可比,还能跨平台转移。

Comments 8 pages, 9 figures, 3 tables, Accepted at IROS 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.04122 2026-07-23 cs.CV 版本更新

Contour Errors: Ego-Centric Matching for 3D Multi-Object Tracking Performance Evaluation

轮廓误差:用于 3D 多目标跟踪性能评估的自我中心匹配

Sharang Kaul, Simon Bultmann, Mario Berk, Abhinav Valada

机构 * CARIAD SE - Volkswagen Group(大众集团CARIAD分公司) Department of Computer Science, University of Freiburg(弗赖堡大学计算机科学系)

AI总结 研究自动驾驶 3D 多目标跟踪开环性能评估问题,提出基于豪斯多夫型推理的轮廓误差(CE)自我中心匹配标准,相比现有方法,CE 能提供分级方向敏感性,评估显示其可提高匹配率,是改善开环感知评估的主要因素。

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

详情

展开后加载摘要…

URL PDF HTML 收藏