Safety-Aligned 3D Object Detection: Single-Vehicle, Cooperative, and End-to-End Perspectives

• URL: http://arxiv.org/abs/2604.03325v1
• Date: Thu, 02 Apr 2026 21:18:01 GMT
• Title: Safety-Aligned 3D Object Detection: Single-Vehicle, Cooperative, and End-to-End Perspectives
• Abstract summary: Perception plays a central role in connected and autonomous vehicles (CAVs)<n>Standard training objectives and evaluation benchmarks treat all perception errors equally, even though only a subset is safety-critical.<n>In this paper, we investigate safety-aligned evaluation and optimization for 3D object detection that explicitly characterize high-impact errors.
• Score: 30.026860273054975
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Perception plays a central role in connected and autonomous vehicles (CAVs), underpinning not only conventional modular driving stacks, but also cooperative perception systems and recent end-to-end driving models. While deep learning has greatly improved perception performance, its statistical nature makes perfect predictions difficult to attain. Meanwhile, standard training objectives and evaluation benchmarks treat all perception errors equally, even though only a subset is safety-critical. In this paper, we investigate safety-aligned evaluation and optimization for 3D object detection that explicitly characterize high-impact errors. Building on our previously proposed safety-oriented metric, NDS-USC, and safety-aware loss function, EC-IoU, we make three contributions. First, we present an expanded study of single-vehicle 3D object detection models across diverse neural network architectures and sensing modalities, showing that gains under standard metrics such as mAP and NDS may not translate to safety-oriented criteria represented by NDS-USC. With EC-IoU, we reaffirm the benefit of safety-aware fine-tuning for improving safety-critical detection performance. Second, we conduct an ego-centric, safety-oriented evaluation of AV-infrastructure cooperative object detection models, underscoring its superiority over vehicle-only models and demonstrating a safety impact analysis that illustrates the potential contribution of cooperative models to "Vision Zero." Third, we integrate EC-IoU into SparseDrive and show that safety-aware perception hardening can reduce collision rate by nearly 30% and improve system-level safety directly in an end-to-end perception-to-planning framework. Overall, our results indicate that safety-aligned perception evaluation and optimization offer a practical path toward enhancing CAV safety across single-vehicle, cooperative, and end-to-end autonomy settings.

Related papers

• From Leaderboard to Deployment: Code Quality Challenges in AV Perception Repositories [4.603321798937855]
  This study systematically analyzed 178 unique models from the KITTI and NuScenes 3D Object Detection leaderboards.<n>Our findings revealed that only 7.3% of the studied repositories meet basic production-readiness criteria.<n>The adoption of Continuous Integration/Continuous Deployment pipelines was correlated with better code maintainability.
  arXiv  Detail & Related papers  (2026-03-02T18:54:28Z)
• EPSM: A Novel Metric to Evaluate the Safety of Environmental Perception in Autonomous Driving [0.5314069314483559]
  It is important to evaluate not only the overall performance of perception systems, but also their safety.<n>We introduce a novel safety metric for jointly evaluating the most critical perception tasks, object and lane detection.<n>Our proposed framework integrates a new, lightweight object safety metric that quantifies the potential risk associated with object detection errors.
  arXiv  Detail & Related papers  (2025-12-17T08:46:49Z)
• Think-Reflect-Revise: A Policy-Guided Reflective Framework for Safety Alignment in Large Vision Language Models [58.17589701432514]
  Think-Reflect-Revise (TRR) is a training framework designed to enhance the safety alignment of Large Vision Language Models (LVLMs)<n>We first build a Reflective Safety Reasoning (ReSafe) dataset with 5,000 examples that follow a think-reflect-revise process.<n>We then fine-tune the target model using the ReSafe dataset to initialize reflective behavior, and finally reinforce policy-guided reflection through reinforcement learning.
  arXiv  Detail & Related papers  (2025-12-08T03:46:03Z)
• SafeR-CLIP: Mitigating NSFW Content in Vision-Language Models While Preserving Pre-Trained Knowledge [51.634837361795434]
  SaFeR-CLIP reconciles safety and performance, recovering up to 8.0% in zero-shot accuracy over prior methods.<n>We also contribute NSFW-Caps, a new benchmark of 1,000 highly-aligned pairs for testing safety under distributional shift.
  arXiv  Detail & Related papers  (2025-11-20T19:00:15Z)
• Real-time Accident Anticipation for Autonomous Driving Through Monocular Depth-Enhanced 3D Modeling [18.071748815365005]
  We introduce an innovative framework, AccNet, which significantly advances the prediction capabilities beyond the current state-of-the-art (SOTA) 2D-based methods. We propose the Binary Adaptive Loss for Early Anticipation (BA-LEA) to address the prevalent challenge of skewed data distribution in traffic accident datasets.
  arXiv  Detail & Related papers  (2024-09-02T13:46:25Z)
• What Makes and Breaks Safety Fine-tuning? A Mechanistic Study [64.9691741899956]
  Safety fine-tuning helps align Large Language Models (LLMs) with human preferences for their safe deployment. We design a synthetic data generation framework that captures salient aspects of an unsafe input. Using this, we investigate three well-known safety fine-tuning methods.
  arXiv  Detail & Related papers  (2024-07-14T16:12:57Z)
• Towards Stable 3D Object Detection [64.49059005467817]
  Stability Index (SI) is a new metric that can comprehensively evaluate the stability of 3D detectors in terms of confidence, box localization, extent, and heading. To help models improve their stability, we introduce a general and effective training strategy, called Prediction Consistency Learning (PCL) PCL essentially encourages the prediction consistency of the same objects under different timestamps and augmentations, leading to enhanced detection stability.
  arXiv  Detail & Related papers  (2024-07-05T07:17:58Z)
• EC-IoU: Orienting Safety for Object Detectors via Ego-Centric Intersection-over-Union [7.355977594790584]
  Ego-Centric Intersection-over-Union (EC-IoU) addresses the limitation of the standard IoU measure in characterizing safety-related performance for object detectors in navigating contexts.<n>We propose a weighting mechanism to refine IoU, allowing it to assign a higher score to a prediction that covers closer points of a ground-truth object from the ego agent's perspective.<n>The proposed EC-IoU measure can be used in typical evaluation processes to select object detectors with better safety-related performance for downstream tasks.
  arXiv  Detail & Related papers  (2024-03-20T16:25:49Z)
• DRUformer: Enhancing the driving scene Important object detection with driving relationship self-understanding [50.81809690183755]
  Traffic accidents frequently lead to fatal injuries, contributing to over 50 million deaths until 2023. Previous research primarily assessed the importance of individual participants, treating them as independent entities. We introduce Driving scene Relationship self-Understanding transformer (DRUformer) to enhance the important object detection task.
  arXiv  Detail & Related papers  (2023-11-11T07:26:47Z)
• A Counterfactual Safety Margin Perspective on the Scoring of Autonomous Vehicles' Riskiness [52.27309191283943]
  This paper presents a data-driven framework for assessing the risk of different AVs' behaviors. We propose the notion of counterfactual safety margin, which represents the minimum deviation from nominal behavior that could cause a collision.
  arXiv  Detail & Related papers  (2023-08-02T09:48:08Z)
• Evaluating Model-free Reinforcement Learning toward Safety-critical Tasks [70.76757529955577]
  This paper revisits prior work in this scope from the perspective of state-wise safe RL. We propose Unrolling Safety Layer (USL), a joint method that combines safety optimization and safety projection. To facilitate further research in this area, we reproduce related algorithms in a unified pipeline and incorporate them into SafeRL-Kit.
  arXiv  Detail & Related papers  (2022-12-12T06:30:17Z)
• USC: Uncompromising Spatial Constraints for Safety-Oriented 3D Object Detectors in Autonomous Driving [7.355977594790584]
  We consider the safety-oriented performance of 3D object detectors in autonomous driving contexts.<n>We present uncompromising spatial constraints (USC), which characterize a simple yet important localization requirement.<n>We incorporate the quantitative measures into common loss functions to enable safety-oriented fine-tuning for existing models.
  arXiv  Detail & Related papers  (2022-09-21T14:03:08Z)
• Neural Network Repair with Reachability Analysis [10.384532888747993]
  Safety is a critical concern for the next generation of autonomy that is likely to rely heavily on deep neural networks for perception and control. This research proposes a framework to repair unsafe DNNs in safety-critical systems with reachability analysis.
  arXiv  Detail & Related papers  (2021-08-09T17:56:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.