Domain generalization (DG) enables generalizing a learning machine from multiple seen source domains to an unseen target one. The general objective of DG methods is to learn semantic representations that are independent of domain labels, which is theoretically sound but empirically challenged due to the complex mixture of common and domain-specific factors. Although disentangling the representations into two disjoint parts has been gaining momentum in DG, the strong presumption over the data limits its efficacy in many real-world scenarios. In this paper, we propose Mix and Reason (\mire), a new DG framework that learns semantic representations via enforcing the structural invariance of semantic topology. \mire\ consists of two key components, namely, Category-aware Data Mixing (CDM) and Adaptive Semantic Topology Refinement (ASTR). CDM mixes two images from different domains in virtue of activation maps generated by two complementary classification losses, making the classifier focus on the representations of semantic objects. ASTR introduces relation graphs to represent semantic topology, which is progressively refined via the interactions between local feature aggregation and global cross-domain relational reasoning. Experiments on multiple DG benchmarks validate the effectiveness and robustness of the proposed \mire.
Unified Modeling of Lane and Lane Topology for Driving Scene Reasoning
School of Artificial Intelligence, Beihang University（北京航空航天大学人工智能学院）；Zhongguancun Academy, Beijing, China（中关村学院，北京，中国）；Hangzhou International Innovation Institute, Beihang University（杭州国际创新院，北京航空航天大学）；Meituan, Beijing, China（美团，北京，中国）
Beijing Institute of Technology（北京理工大学）；MEGVII Technology（旷视科技）；University of Science and Technology of China（中国科学技术大学）；SKL-IOTSC, University of Macau（澳门大学智慧城市物联网国家重点实验室）
Topology reasoning aims to comprehensively understand road scenes and present drivable routes in autonomous driving. It requires detecting road centerlines (lane) and traffic elements, further reasoning their topology relationship, i.e., lane-lane topology, and lane-traffic topology. In this work, we first present that the topology score relies heavily on detection performance on lane and traffic elements. Therefore, we introduce a powerful 3D lane detector and an improved 2D traffic element detector to extend the upper limit of topology performance. Further, we propose TopoMLP, a simple yet h…
RATopo: Improving Lane Topology Reasoning via Redundancy Assignment
School of Artificial Intelligence Beihang University Beijing China（人工智能学院 北航 北京 中国）；Zhongguancun Academy Beijing China（中关村学院 北京 中国）；Hangzhou International Innovation Institute Beihang University Hangzhou China（杭州国际创新研究院 北航 杭州 中国）；Beihang University（北航）；Zhongguancun Academy（中关村学院）；University of Macau（澳门大学）
TopoLogic: An Interpretable Pipeline for Lane Topology Reasoning on Driving Scenes
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Hangzhou Dianzi University（杭州电子科技大学）
Coherent Online Road Topology Estimation and Reasoning with Standard-Definition Maps
CARIAD SE and the Technical University Munich（CARIAD SE 和慕尼黑技术大学）；Technical University Munich（慕尼黑技术大学）；Munich Institute of Robotics and Machine Intelligence (MIRMI)（慕尼黑机器人与机器智能研究所）；Center for Robotics, University of Bonn（波恩大学机器人中心）；Lamarr Institute for Machine Learning and Artificial Intelligence（拉马尔机器学习与人工智能研究所）
TopoHR: Hierarchical Centerline Representation for Cyclic Topology Reasoning in Driving Scenes with Point-to-Instance Relations
Institute of Intelligent Machines, HFIPS, Chinese Academy of Sciences（智能机器研究所，HFIPS，中国科学院）；University of Science and Technology of China（中国科学技术大学）；NullMax；Westlake University（西湖大学）
TopoPoint: Enhance Topology Reasoning via Endpoint Detection in Autonomous Driving
Institute of Computing Technology, Chinese Academy of Science（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Shanghai AI Lab（上海人工智能实验室）；Shanghai Innovation Institute（上海创新研究院）
A Concise Survey on Lane Topology Reasoning for HD Mapping
NavInfo Co., Ltd.（四维图新有限公司）；Autohome Inc.（汽车之家）；Baidu Inc.（百度公司）；Xidian University of Technology（西安工业大学）；Hefei University of Technology（合肥工业大学）；Renmin University of China（中国人民大学）
SEPT: Standard-Definition Map Enhanced Scene Perception and Topology Reasoning for Autonomous Driving
Department of Electronic and Computer Engineering, 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）；Zhuoyu Technology（卓予科技）；State Key Laboratory for Novel Software Technology, Nanjing University（新型软件技术国家重点实验室，南京大学）
OpenDriveLab（开放驾驶实验室）；Shanghai AI Lab（上海人工智能实验室）；Huawei（华为）；Fudan University（复旦大学）；The University of Hong Kong（香港大学）；Shanghai Jiao Tong University（上海交通大学）
Chameleon: Fast-slow Neuro-symbolic Lane Topology Extraction
Institute for AI Industry Research (AIR), Tsinghua University（清华大学人工智能产业研究院（AIR））；Bosch Corporate Research, China（博世中国企业研究中心）；Institute for Interdisciplinary Information Sciences(IIIS), Tsinghua University（清华大学交叉信息研究院）
