Core Insights - The article discusses advancements in trajectory prediction using a generative active learning framework called GALTraj, which applies controllable diffusion models to address long-tail issues in data [1][2]. Group 1: GALTraj Framework - GALTraj is the first framework to apply generative active learning to trajectory prediction tasks, enhancing long-tail learning without modifying the model structure [2]. - The framework employs a tail-aware generation method that differentiates the diffusion guidance for tail, head, and related agents, producing realistic and diverse scenarios while preserving tail characteristics [2][3]. Group 2: Experimental Results - In experiments on WOMD and Argoverse2 datasets, GALTraj significantly improved long-tail sample prediction performance, reducing the long-tail metric FPR₅ by 47.6% (from 0.42 to 0.22) and overall prediction error minFDE₆ by 14.7% (from 0.654 to 0.558) [1][6]. - The results indicate that GALTraj outperforms traditional methods across various metrics, showcasing its effectiveness in enhancing prediction accuracy for rare scenarios [7][8]. Group 3: TopoLiDM Framework - The article also highlights the TopoLiDM framework developed by Shanghai Jiao Tong University and Twente University, which integrates topology-aware diffusion models for high-fidelity LiDAR point cloud generation [13][15]. - TopoLiDM achieved a 22.6% reduction in the Fréchet Range Image Distance (FRID) and a 9.2% reduction in Minimum Matching Distance (MMD) on the KITTI-360 dataset while maintaining a real-time generation speed of 1.68 samples per second [13][15]. Group 4: FastDriveVLA Framework - FastDriveVLA, developed by Peking University and Xiaopeng Motors, introduces a reconstruction-based visual token pruning framework that maintains 99.1% trajectory accuracy with a 50% pruning rate and reduces collision rates by 2.7% [21][22]. - The framework employs a novel adversarial foreground-background reconstruction strategy to enhance the identification of valuable tokens, achieving state-of-the-art performance on the nuScenes open-loop planning benchmark [27][28]. Group 5: PLA Framework - The article presents a unified Perception-Language-Action (PLA) framework proposed by TUM, which integrates multi-sensor fusion and GPT-4.1 enhanced visual-language-action reasoning for adaptive autonomous driving [34][35]. - The framework demonstrated a mean absolute error (MAE) of 0.39 m/s in speed prediction and an average displacement error (ADE) of 1.013 meters in trajectory tracking within urban intersection scenarios [42].

Core Viewpoint - The article emphasizes the convergence of autonomous driving technology, highlighting the shift from numerous diverse approaches to a more unified model, which indicates higher technical barriers in the industry [1] Group 1 - The industry is moving towards a unified solution with models like one model, VLM, and VLA, suggesting a reduction in the need for numerous algorithm engineers [1] - The article encourages the establishment of a large community to support industry professionals, facilitating growth and collaboration among peers [1] - A new job-related community is being launched to discuss industry trends, company developments, product research, and job opportunities [1]

小马智行表示，运营时段充分满足市民包括早晚高峰在内的日常通勤及休闲出行需求，同时运营线路覆 盖浦东新区人民政府、啦啦宝都购物中心、世纪公园、浦东足球场地铁站等核心商圈、地标场所及热门 地铁站，有效连接区域内工作、生活、休闲三大场景，让自动驾驶技术真正服务于市民的日常出行需 求。 责任编辑：王翔 消息称，小马智行联合锦江出租率先在浦东金桥和花木地区推出了面向公众常态化运营的Robotaxi服 务。每周一至周五早7：30至晚9：30，市民通过"小马智行"手机App或小程序，均可呼叫Robotaxi作为 出行交通工具。运营线路覆盖浦东新区人民政府、啦啦宝都购物中心、世纪公园、浦东足球场地铁站等 核心商圈、地标场所及热门地铁站。 8月4日消息，小马智行宣布，8月1日起，小马智行在上海浦东推出了面向公众常态化运营的Robotaxi服 务。 ...

作者 | 王小迪MLE 编辑 | 自动驾驶之心 原文链接： https://www.zhihu.com/question/279973696/answer/3535722816 点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近15个 方向 学习 路线 >>自动驾驶前沿信息获取 → 自动驾驶之心知识星球 本文只做学术分享，如有侵权，联系删文 最近end2end风头正盛，BEV已成标准范式，但planning进展仍然焦灼。究其原因，interaction modelling是planning的深水区，涉及博弈和不确定性建模，监督学习仍然 不能很好得描述此类问题。这次报告以interaction的建模和求解为切口，分析了近些年常用的框架范式，比如将ego trajectory和agent trajectory的关系加入loss function或 constraint中，ego/agent trajectory从lane routing或neural network生成等。 - (We already have) Reactive: surrounding agents influenc ...

微信扫码添加小助理邀请进群， 备注自驾+昵称+求职 ； 最近和很多准备校招的小伙伴接触，发现大家在学校学习的东西和工作的差距越来越大。有不少工作多年 的小伙伴表示也在看机会，感知转大模型、世界模型，传统规控想转具身。但却不知道业内实际在做什 么，导致秋招的时候没有什么优势...... 峰哥一直在鼓励大家坚持、多和其他人交流，但归根结底个人的力量是有限的。我们希望共建一个大的社 群和大家一起成长，真正能够帮助到一些有需要的小伙伴，成为一个汇集全行业人才的综合型平台，真正 做一个链接学校和公司的桥梁。所以我们也开始正式运营求职与行业相关的社群。社群内部主要讨论相关 产业、公司、产品研发、求职与跳槽相关内容。如果您想结交更多同行业的朋友，第一时间了解产业。欢 迎加入我们！ ...

Core Viewpoint - The article discusses a new detection framework called Butter, designed to improve target detection in autonomous driving scenarios by addressing the challenges of multi-scale semantic information modeling and enhancing detection robustness and deployment efficiency [3][11]. Group 1: Framework Innovations - Butter introduces two core innovations in the Neck layer: the Frequency Consistency Enhancement Module (FAFCE) and the Progressive Hierarchical Feature Fusion Network (PHFFNet) [3][15]. - FAFCE enhances boundary resolution by integrating high-frequency detail enhancement with low-frequency noise suppression, while PHFFNet progressively fuses semantic information to strengthen multi-scale feature representation [3][15]. Group 2: Performance Metrics - Butter outperforms existing state-of-the-art (SOTA) methods in detection accuracy with significantly lower parameter counts, achieving a mean Average Precision (mAP@50) of 94.4% on the KITTI dataset, surpassing the previous best by 1.2 percentage points while using only about one-third of the computational load [32][34]. - On the BDD100K and Cityscapes datasets, Butter achieved mAP@50 scores of 53.7% and 53.2%, respectively, demonstrating superior performance compared to other lightweight models, particularly with a 1.6 percentage point improvement on Cityscapes [32][34]. Group 3: Structural Challenges - Existing Neck structures often face issues such as frequency aliasing and rigid fusion processes, which compromise feature expression and detection accuracy, particularly for small targets in complex environments [9][10]. - Butter's design addresses these structural bottlenecks by decoupling frequency modeling and multi-scale fusion, achieving a balance between accuracy and efficiency [11][12]. Group 4: Methodology Overview - The Butter framework begins with a 640×640 monocular image, extracting initial features through a lightweight Backbone module, followed by refinement through various lightweight blocks before entering the Neck module [16][17]. - The model employs a four-output head in the Head layer to generate final detection results, including class labels, confidence scores, and bounding boxes [16][17]. Group 5: Feature Fusion Techniques - FAFCE enhances feature fusion accuracy and robustness by employing high-frequency amplification and low-frequency damping mechanisms, which improve the consistency and precision of multi-scale feature integration [20][27]. - PHFFNet implements a hierarchical fusion strategy that alleviates semantic discrepancies between non-adjacent layers, significantly enhancing detection accuracy and alignment in scenarios requiring precise boundary detection [29][30].

WeRide plans to ramp up spending, even as it continues to post losses, as the Chinese autonomous driving company bets that its ambitious international expansion plans will eventually see it reach profitability https://t.co/OzOplQ7iFe ...

Financial Data and Key Metrics Changes - Total revenue for Q2 2025 reached RMB 127.2 million, a 60.8% increase year over year, driven by strong growth in both product and service segments [20] - Product revenue surged by 309.6% year over year to RMB 59.8 million, with robotaxi revenue hitting a record high of RMB 45.9 million, up 836.7% year over year, contributing 36.1% to total revenue [21] - Group level gross profit increased by 40.6% to RMB 35.7 million, with a gross margin of 28.1% [23] - Net loss decreased by 1.7% to RMB 406.4 million, while on a non-IFRS basis, net loss increased to RMB 306.6 million due to ongoing R&D investments [26] Business Line Data and Key Metrics Changes - Robotaxi revenue accounted for 36% of total revenue, indicating a strong momentum that is expected to continue [36] - Service revenue grew by 4.3% year over year to RMB 67.4 million, supported by intelligent data services and L4 operational support [22] Market Data and Key Metrics Changes - The company operates the largest public commercial robotaxi fleet outside the US and China, with significant expansion in the Middle East, particularly in Abu Dhabi and Saudi Arabia [11][12] - The robotaxi fleet in Dubai has tripled in size, covering approximately 50% of core areas, with plans to further scale the fleet [12] Company Strategy and Development Direction - The company aims to transform future mobility through safe and accessible driverless solutions, with a focus on expanding global robotaxi operations [10] - A multi-product strategy is employed, leveraging a universal platform that supports various urban mobility applications, enhancing data collection and operational flexibility [30][32] - The Middle East is identified as a strategic priority for growth, with plans to expand robotaxi services to 15 additional cities [13] Management's Comments on Operating Environment and Future Outlook - Management expressed confidence in the company's ability to convert global potential into long-term value, citing successful partnerships and regulatory momentum [26] - The company is optimistic about the future, with plans to scale operations and enhance user experience through advanced technology [49] Other Important Information - The company has received autonomous driving permits in six countries, demonstrating its technology's compliance with high global safety standards [22][19] - The newly launched HPC 3.0 computing platform is expected to cut costs by 50% and enhance the performance of autonomous vehicles [17][48] Q&A Session Summary Question: Could you elaborate on the company's multi-product strategy and the relationship between Robotaxi and other products? - The company emphasizes a multi-product strategy that allows for data sharing across different autonomous vehicle types, enhancing system improvement and market adaptability [30][32] Question: Is the revenue contribution from the Robotaxi business expected to sustain in the coming quarters? - Management believes the strong momentum in Robotaxi revenue will continue, driven by fleet expansion and operational scaling in key markets [36][37] Question: When will the HPC 3.0 platform be deployed in the next generation robotaxi? - The HPC 3.0 platform is already in use, with the Robotaxi GXR being the first mass-produced L4 autonomous vehicle utilizing this technology [45][48] Question: How many permits are currently in the pipeline and what are the expansion targets for Robotaxi? - The company has over 1,300 autonomous vehicles globally, with plans to add hundreds more by the end of the year, focusing on markets with strong unit economics [54][56] Question: What challenges does the company face in deploying robotaxi services in new markets? - Challenges include adapting technology to local conditions and navigating regulatory frameworks, but the company has built a strong foundation for successful deployment [90][92] Question: How does the company view the competitive landscape of robotaxi operations? - The company is confident in its competitive edge due to its extensive operational experience, safety track record, and advanced technology [75][78] Question: How does the company leverage new AI models for future technology trends? - The company is integrating advanced AI models into its autonomous driving systems, enhancing training and operational capabilities [81][84]

NEW YORK, July 31, 2025 (GLOBE NEWSWIRE) -- WeRide Inc. ("WeRide" or the "Company") (Nasdaq: WRD), a global leader in autonomous driving technology, today announced its unaudited financial results for the three months ended June 30, 2025. Recent Highlights Redefining Autonomous Driving with Leading Technology Breakthroughs High-Performance Computing (HPC) 3.0 Platform Powered by NVIDIA DRIVE AGX Thor Chips Partnership with Chery on New-Generation Robotaxi to Revolutionize Urban Mobility Unparalleled Robotax ...

本文只做学术分享，如有侵权，联系删文 写在前面--先聊聊为啥写这篇文章 笔者这段时间阅读了来自ICCV2025的论文 DONUT: A Decoder-Only Model for Trajectory Prediction 作者 | Sakura 编辑 | 自动驾驶之心 原文链接： https://zhuanlan.zhihu.com/p/1933901730589962575 点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近15个 方向 学习 路线 >>自动驾驶前沿信息获取 → 自动驾驶之心知识星球 这篇论文以qcnet为baseline，基于 decoder-only架构配合overprediction策略 ，在argoversev2上取得了SOTA 联想到之前笔者所阅读的论文SmartRefine，该论文也是基于Qcnet的基础上对refine部分进行改进，也在argoverse v2上取得了SOTA； 因此，本着学习的态度，笔者想 在此简单总结这三篇论文 ； Query-Centric Trajectory Prediction--CVPR 2023 SmartRefin ...