自动驾驶 - arXivDaily 专题

1608.02051 2016-08-09 cs.CV 57%

Compressive Change Retrieval for Moving Object Detection

Tomoya Murase, Kanji Tanaka

专题命中感知：autonomous driving(abstract)；分类 cs.CV

Comments 6 pages, 6 figures, Draft of a paper submitted to an International Conference

1604.02469 2016-04-12 cs.CV 57%

Image segmentation of cross-country scenes captured in IR spectrum

Artem Lenskiy

专题命中感知：self-driving(abstract)；分类 cs.CV

Comments Corrected version of the chapter published in Advances in Robotics and Virtual Reality, Volume 26 of the series Intelligent Systems Reference Library pp 227-247

1601.05535 2016-01-22 cs.CV 57%

On the Diagnostic of Road Pathway Visibility

Pierre Charbonnier, Jean-Philippe Tarel, Francois Goulette

专题命中感知：LiDAR(abstract)；分类 cs.CV

Comments in Transport Research Arena Europe, 2010, Bruxelles, France. 2010

1511.00513 2015-11-03 cs.CV 57%

Pixel-wise Segmentation of Street with Neural Networks

Sebastian Bittel, Vitali Kaiser, Marvin Teichmann, Martin Thoma

专题命中感知：self-driving(abstract)；分类 cs.CV

1506.04723 2015-07-30 cs.CV 57%

Layered Interpretation of Street View Images

Ming-Yu Liu, Shuoxin Lin, Srikumar Ramalingam, Oncel Tuzel

专题命中感知：autonomous driving(abstract)；分类 cs.CV

Comments The paper will be presented in the 2015 Robotics: Science and Systems Conference (RSS)

1309.4385 2015-06-17 physics.optics cs.CV 57%

Photon counting compressive depth mapping

Gregory A. Howland, Daniel J. Lum, Matthew R. Ware, John C. Howell

专题命中感知：LiDAR(abstract)；分类 cs.CV

Comments 16 pages, 8 figures

1503.06643 2015-03-24 cs.CV 57%

A novel pLSA based Traffic Signs Classification System

Mrinal Haloi

专题命中感知：autonomous driving(abstract)；分类 cs.CV

Comments APMediaCast-2015, Bali, Indonesia

1412.3506 2014-12-19 cs.CV 57%

Road Detection by One-Class Color Classification: Dataset and Experiments

Jose M. Alvarez, Theo Gevers, Antonio M. Lopez

专题命中感知：autonomous driving(abstract)；分类 cs.CV

Comments 10 pages

1401.5657 2014-01-23 cs.RO 57%

Enhancing Mobile Object Classification Using Geo-referenced Maps and Evidential Grids

Marek Kurdej, Julien Moras, Véronique Cherfaoui, Philippe Bonnifait

专题命中感知：occupancy(abstract)；分类 cs.RO

Comments 6 pp. arXiv admin note: substantial text overlap with arXiv:1207.1016

Journal ref IEEE/RSJ International Conference on Intelligent Robots and Systems. 5th Workshop on Planning, Perception and Navigation for Intelligent Vehicles, Tokyo : Japan (2013)

1301.0551 2013-01-07 cs.LG cs.RO stat.ML 57%

Learning Hierarchical Object Maps Of Non-Stationary Environments with mobile robots

Dragomir Anguelov, Rahul Biswas, Daphne Koller, Benson Limketkai, Sebastian Thrun

专题命中感知：occupancy(abstract)；分类 cs.RO

Comments Appears in Proceedings of the Eighteenth Conference on Uncertainty in Artificial Intelligence (UAI2002)

1204.0067 2012-04-03 cs.RO 57%

Estimating Rigid Transformation Between Two Range Maps Using Expectation Maximization Algorithm

Shuqing Zeng

专题命中感知：LiDAR(abstract)；分类 cs.RO

1012.4776 2010-12-22 cs.AI 57%

Automatic Estimation of the Exposure to Lateral Collision in Signalized Intersections using Video Sensors

Nicolas Saunier, Sophie Midenet

专题命中感知：occupancy(abstract)；分类 cs.AI

2301.11324 2023-01-27 cs.LG 56%

Certified Interpretability Robustness for Class Activation Mapping

Alex Gu, Tsui-Wei Weng, Pin-Yu Chen, Sijia Liu, Luca Daniel

专题命中感知：autonomous driving(abstract,comments)

Comments 13 pages, 5 figures. Accepted to Machine Learning for Autonomous Driving Workshop at NeurIPS 2020

1312.5224 2013-12-19 stat.AP 56%

The passive and active periods for the intermittent use of an active sensor to detect an evasive target

Niels Bache

专题命中感知：LiDAR(abstract,comments)

Comments Key words: Active sensor. Sonar. Radar. Lidar. Evasive target. Detection. Missed opportunities. Detection width. Forestalling. Counter detection

2608.14917 2026-08-18 cs.CE 新提交 50%

Ground-Truth-Aware Stress Testing of a Closed-Loop Smart-Building Digital Twin Under Sensor Drift and Missing Data

传感器漂移与缺失数据下闭环智能建筑数字孪生的真值感知压力测试

Lyes Saad Saoud

专题命中感知：occupancy(abstract)

AI总结本文提出真值感知仿真框架，针对传感器漂移与缺失数据场景测试智能建筑数字孪生的闭环控制，发现传感器误差、决策不一致与结果退化相关但有区别，贡献为方法论层面。

2608.13962 2026-08-17 cs.AR 新提交 50%

MoE Expert Execution in Disaggregated LLM Serving with a High-Bandwidth ReRAM Near-Memory Architecture

采用高带宽ReRAM近内存架构的分散式大语言模型（LLM）服务中的MoE专家执行

Kunming Shao, Ming Zeng, Xin Yuan, Binbin Liao, Yangming Zhang, Wei Wang, Tim Kwang-Ting Cheng, Chi-Ying Tsui

专题命中感知：occupancy(abstract)

AI总结本文针对分散式LLM服务中MoE专家执行的资源效率问题，提出ReRAM近内存架构，通过优化池化、放置与取数策略提升资源占用率，在实测中大幅降低延迟与能耗。

2608.11431 2026-08-13 cs.LG 新提交 50%

AutoGrable: What Is a Good Graph for a Table?

AutoGrable：什么是适合表格的图？

Tamara Cucumides, Floris Geerts

机构 * University of Antwerp（安特卫普大学）

专题命中感知：occupancy(abstract)

AI总结 AutoGrable是一种无需训练图模型的表格到图构造方法，通过标签对齐风险打分选择最优列子集，在受控和真实任务中性能优于对比方法，还可在图无用时弃权构建。

Comments 28 pages, 4 figures

2608.10402 2026-08-12 cs.LG cs.DC 新提交 50%

TideRL: Boosting Agentic RL Goodput with Readiness-Aware Scheduling

TideRL：通过就绪感知调度提升智能体强化学习的有效吞吐量

Yanyu Ren, Xizheng Wang, Xiao Liu, Bowen Lv, Hanchen Zhang, Shudan Zhang, Hanyu Lai, Shuai Wang, Li Chen, Dan Li, Jie Tang

机构 * Tsinghua University（清华大学）； Zhongguancun Laboratory（中关村实验室）

专题命中感知：occupancy(abstract)

AI总结 TideRL是一种就绪感知弹性RL系统，通过CTB、RA²P和ERS技术，在多轮智能体工作负载上大幅提升RL训练有效吞吐量，同时优化KV缓存命中率、训练时间和等待时间。

2608.10204 2026-08-12 cs.LG 新提交 50%

Boundary-Seeking Policy Gradient for Safe Reinforcement Learning

面向安全强化学习的边界搜寻策略梯度算法

Chenhua Fan, Jiahui Zhu, Yuhang Zhang, Honghao Wei

机构 * School of EECS, Washington State University（华盛顿州立大学电子工程与计算机科学学院）

专题命中感知：occupancy(abstract)

AI总结本文提出BSPG算法，针对安全强化学习中标准梯度法易收敛到可行域内部的问题，结合切向与法向分量，在Safety-Gymnasium导航任务中实现更高奖励与更紧密的边界跟踪。

Comments CDC 2026

2608.08425 2026-08-11 cs.NI cs.DC 新提交 50%

PSP: Low-Overhead Packet-Level Load Balancing for Stale-State and Bandwidth-Asymmetric Networks

PSP：面向 stale 状态与带宽非对称网络的低开销报文级负载均衡

Jiaqi Liu, Chunyang Zhang, Heng Pan, Yanbiao Li

专题命中感知：occupancy(abstract)

AI总结针对数据中心网络微突发与高并发问题，提出概率状态比例（PSP）报文级负载均衡算法，其性能优于 JSQ、随机调度，与 Top-k 相当且硬件成本更低，实现性能、稳定性与开销的平衡。

Comments 12 pages, 10 figures, 5 tables. Accepted by IEEE LCN 2026

2608.07725 2026-08-11 cs.LG 新提交 50%

Finite Constant Frontiers and Auditable Regret Certificates for Average-Reward Reinforcement Learning

平均奖励强化学习的有限常数前沿与可审计后悔证书

Ibne Farabi Shihab, Abu Sa-Adat Mohamed Moon-Im Al Ahsan, Md Najmus Swaqeeb

机构 * Iowa State University（爱荷华州立大学）； BRAC University（BRAC大学）

专题命中感知：occupancy(abstract)

AI总结本文针对平均奖励强化学习，提出感知常数的比较协议，推导通信MDP的有限下界证书，改进已发表系数，给出跨度约束乐观学习者的可审计组合规则，完成相关形式化与诊断测试。

2605.25316 2026-08-11 eess.SP 版本更新 50%

An Extended Object Poisson Multi-Bernoulli Filter with Zero-Inflated Poisson Measurement Model Using Belief Propagation

基于零膨胀泊松模型和置信传播的扩展目标PMB滤波器

Xueqi Qiu, Yuxuan Xia, Hyowon Kim, Chaoqun Yang

专题命中感知：LiDAR(abstract)

AI总结提出一种利用粒子置信传播高效实现零膨胀泊松模型下扩展目标泊松多伯努利滤波器的方法，通过因子图分解数据关联，提高估计精度和运行速度。

Comments 24 pages, 5 figures

2608.05634 2026-08-07 cs.MM 新提交 50%

MEC-Patch: Visible-Infrared Cross-Modal Adversarial Attack Driven by Intrinsic Material Emissivity Laws

MEC-Patch：基于本征材料发射率定律的可见-红外跨模态对抗攻击

Zhixiang Huang, Xinbo Nie, Wenxuan Wang, Lu Yang, Xin Li, Xuelin Qian, Peng Wang

专题命中感知：autonomous driving(abstract)

AI总结本文提出基于斯特藩-玻尔兹曼定律的MEC-Patch跨模态对抗攻击框架，结合NSGA-II与DAR策略，可欺骗多模态检测器并提升环境鲁棒性，为多模态感知系统安全评估提供新思路。

Comments Accepted by the 34th ACM International Conference on Multimedia (MM '26). Includes supplementary material

2608.04852 2026-08-06 cs.ET 新提交 50%

Toward Blockage-Resilient 6G-V2X Connectivity: Semi-Distributed Bandit with Dynamic Arm Set for mmWave HetNets

面向抗阻塞的6G车万物联网（V2X）连接：面向毫米波异构车联网的具有动态臂集的半分布式多臂老虎机算法

Weiqi Chi, Bo Qian, Hanlin Wu, Donghui Li, Haibo Zhou, Manabu Tsukada

专题命中感知：autonomous driving(abstract)

AI总结针对毫米波异构车联网的阻塞与信道波动问题，提出BAND及半分布式S-BAND算法，结合TAK区域与KIF度量，在10%-50%阻塞率下较集中式MAB基线实现34.9%-59.4%的遗憾降低，TAK区域性能优于K-means聚类。

2608.04068 2026-08-06 cs.IT eess.SP math.IT 新提交 50%

Tail-Calibrated Soft-Output GRAND for Finite-Memory Noise-Effect Posteriors

面向有限记忆噪声效应后验的尾端校准软输出GRAND

Behrooz Razeghi

专题命中感知：occupancy(abstract)

AI总结针对有限记忆噪声效应后验，提出尾端校准软输出GRAND算法，实现后验权重等估计并给出相关理论界，可提供逐块APP估计等解码输出。

2607.26665 2026-08-05 cs.ET 版本更新 50%

OpenCSI: Self-Calibration Layer for Heterogeneous Mesh Wireless Sensor Networks

OpenCSI：面向异构网状无线传感器网络的自校准层

Karim Khamaisi, Simon Sigg, Bruno Rodrigues

专题命中感知：occupancy(abstract)

AI总结 OpenCSI是面向异构网状无线传感器网络的自校准层，通过在线学习空房间基线实现WiFi CSI感知模型零样本跨环境迁移，在二元占用任务上F1最高达0.99，无需目标域数据或重训练。

2602.13542 2026-08-04 cs.NI cs.DC eess.SP 交叉投稿 50%

SIDSense: Database-Free TV White Space Sensing for Disaster-Resilient Connectivity

SIDSense：无数据库的电视白空间传感以实现灾害韧性连接

George M. Gichuru, Zoe Aiyanna M. Cayetano

专题命中感知：occupancy(abstract)

AI总结 SIDSense通过无数据库的TVWS传感技术，结合边缘AI和私有5G堆栈，在灾害应急通信中实现高可靠连接和低延迟感知

Comments 7 pages, 1 figure

2607.27371 2026-07-31 eess.SY cs.SY 新提交 50%

Dynamics-matched Physical Reservoir Computing for Undersensed Traffic Prediction

动力学匹配物理储备池计算用于欠感知交通预测

Michael McCreesh, Rohit Gupta, Stephen L. Smith

专题命中感知：autonomous driving(abstract)

AI总结本研究提出将交通网络作为储备池的IIDM-RC方法，用于欠感知交通预测，在预测准确性和训练时间上优于ESNs与LSTM网络。

Comments 16 pages, 4 figures. Has been submitted for journal publication

2607.25114 2026-07-29 eess.SY cs.SY math.OC 新提交 50%

Integrated Eco-Driving and Powertrain Optimization for Hybrid Vehicles in Complex Urban Traffic

复杂城市交通中混合动力汽车的集成生态驾驶与动力总成优化

Milad Hasanzadeh, Amin Kargarian, Mehdi Farasat

专题命中感知：occupancy(abstract)

AI总结针对城市混合动力汽车，提出集成有限时域生态驾驶规划框架，联合优化交通行为与动力总成运行，同时建模多方面驾驶行为及动力总成运行，模拟研究表明该方法能在维持合规驾驶计划时改善能源性能。

2607.23059 2026-07-28 cs.CR 新提交 50%

GLST: Defending Confidence-Driven V2X Collaborative Perception Against Stealthy Multi-Attacker Feature Injection

GLST：抵御针对置信驱动的V2X协同感知的隐秘多攻击者特征注入

Ji He, Ying Wang, Lijie Zheng, Xinghui Zhu, Yulong Shen, Xiaohong Jiang

专题命中感知：driving perception(abstract)

AI总结研究针对置信驱动的V2X协同感知的安全问题，提出伪装良性攻击利用空间置信机制降低检测性能。发现现有基于信任的防御弱点，进而提出GLST通过多视角评估协作者可靠性，实验表明其在多攻击者场景下鲁棒性强，对多种攻击有效。

1. 感知 6065 篇

Compressive Change Retrieval for Moving Object Detection

Image segmentation of cross-country scenes captured in IR spectrum

On the Diagnostic of Road Pathway Visibility

Pixel-wise Segmentation of Street with Neural Networks

Layered Interpretation of Street View Images

Photon counting compressive depth mapping

A novel pLSA based Traffic Signs Classification System

Road Detection by One-Class Color Classification: Dataset and Experiments

Enhancing Mobile Object Classification Using Geo-referenced Maps and Evidential Grids

Learning Hierarchical Object Maps Of Non-Stationary Environments with mobile robots

Estimating Rigid Transformation Between Two Range Maps Using Expectation Maximization Algorithm

Automatic Estimation of the Exposure to Lateral Collision in Signalized Intersections using Video Sensors

Certified Interpretability Robustness for Class Activation Mapping

The passive and active periods for the intermittent use of an active sensor to detect an evasive target

Ground-Truth-Aware Stress Testing of a Closed-Loop Smart-Building Digital Twin Under Sensor Drift and Missing Data

MoE Expert Execution in Disaggregated LLM Serving with a High-Bandwidth ReRAM Near-Memory Architecture

AutoGrable: What Is a Good Graph for a Table?

TideRL: Boosting Agentic RL Goodput with Readiness-Aware Scheduling

Boundary-Seeking Policy Gradient for Safe Reinforcement Learning

PSP: Low-Overhead Packet-Level Load Balancing for Stale-State and Bandwidth-Asymmetric Networks

Finite Constant Frontiers and Auditable Regret Certificates for Average-Reward Reinforcement Learning

An Extended Object Poisson Multi-Bernoulli Filter with Zero-Inflated Poisson Measurement Model Using Belief Propagation

MEC-Patch: Visible-Infrared Cross-Modal Adversarial Attack Driven by Intrinsic Material Emissivity Laws

Toward Blockage-Resilient 6G-V2X Connectivity: Semi-Distributed Bandit with Dynamic Arm Set for mmWave HetNets

Tail-Calibrated Soft-Output GRAND for Finite-Memory Noise-Effect Posteriors

OpenCSI: Self-Calibration Layer for Heterogeneous Mesh Wireless Sensor Networks

SIDSense: Database-Free TV White Space Sensing for Disaster-Resilient Connectivity

Dynamics-matched Physical Reservoir Computing for Undersensed Traffic Prediction

Integrated Eco-Driving and Powertrain Optimization for Hybrid Vehicles in Complex Urban Traffic

GLST: Defending Confidence-Driven V2X Collaborative Perception Against Stealthy Multi-Attacker Feature Injection