mlsanityData to 7 Oct 2026

Towards Driver Behavior Understanding: Weakly-Supervised Risk Perception in Driving Scenes

arXiv 2603.05926 ↗First posted 6 Mar 2026cs.CVcited by 1

Not on the map. Why?

Abstract (arXiv)

Achieving zero-collision mobility remains a key objective for intelligent vehicle systems, which requires understanding driver risk perception-a complex cognitive process shaped by voluntary response of the driver to external stimuli and the attentiveness of surrounding road users towards the ego-vehicle. To support progress in this area, we introduce RAID (Risk Assessment In Driving scenes)-a large-scale dataset specifically curated for research on driver risk perception and contextual risk assessment. RAID comprises 4,691 annotated video clips, covering diverse traffic scenarios with labels for driver's intended maneuver, road topology, risk situations (e.g., crossing pedestrians), driver responses, and pedestrian attentiveness. Leveraging RAID, we propose a weakly supervised risk object identification framework that models the relationship between driver's intended maneuver and responses to identify potential risk sources. Additionally, we analyze the role of pedestrian attention in estimating risk and demonstrate the value of the proposed dataset. Experimental evaluations demonstrate that our method achieves 20.6% and 23.1% performance gains over prior state-of-the-art approaches on the RAID and HDDS datasets, respectively.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 38 of its 60 references that are in the map. First posted on arXiv 6 Mar 2026.

Builds on

38 of 60 references.

References · 38

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  3. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
  4. 2017Mask R-CNNcited by 7,259
  5. 2014Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modelingcited by 2,341
  6. 2017Simple Online and Realtime Tracking with a Deep Association Metriccited by 908
  7. 2018Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networkscited by 687
  8. 2016MOT16: A Benchmark for Multi-Object Trackingcited by 531
  9. 2018Image Inpainting for Irregular Holes Using Partial Convolutionscited by 505
  10. 2016Temporal Convolutional Networks for Action Segmentation and Detectioncited by 504
  11. 2019Tracking without bells and whistlescited by 382
  12. 2015WIDER FACE: A Face Detection Benchmarkcited by 297
26 more
  1. 2019Causal Confusion in Imitation Learningcited by 286
  2. 2019RetinaFace: Single-stage Dense Face Localisation in the Wildcited by 164
  3. 2022DRAMA: Joint Risk Localization and Captioning in Drivingcited by 125
  4. 2019Seamless Scene Segmentationcited by 112
  5. 2017Interpretable Learning for Self-Driving Cars by Visualizing Causal Attentioncited by 106
  6. 2019Learning Actor Relation Graphs for Group Activity Recognitioncited by 83
  7. 2018Autonomous Vehicles that Interact with Pedestrians: A Survey of Theory and Practicecited by 74
  8. 2018Deep Cosine Metric Learning for Person Re-Identificationcited by 73
  9. 2020Spatiotemporal Relationship Reasoning for Pedestrian Intent Predictioncited by 71
  10. 2023Rank2Tell: A Multimodal Driving Dataset for Joint Importance Ranking and Reasoningcited by 63
  11. 2017Repulsion Loss: Detecting Pedestrians in a Crowdcited by 62
  12. 2017Predicting Driver Attention in Critical Situationscited by 61
  13. 2018Occlusion-aware R-CNN: Detecting Pedestrians in a Crowdcited by 58
  14. 2020TITAN: Future Forecast using Action Priorscited by 58
  15. 2021LOKI: Long Term and Key Intentions for Trajectory Predictioncited by 57
  16. 2020Pedestrian Action Anticipation using Contextual Feature Fusion in Stacked RNNscited by 48
  17. 2018Deep Object-Centric Policies for Autonomous Drivingcited by 41
  18. 2020Who Make Drivers Stop? Towards Driver-centric Risk Assessment: Risk Object Identification via Causal Inferencecited by 34
  19. 2017Agent-Centric Risk Assessment: Accident Anticipation and Risky Region Localizationcited by 33
  20. 2019Goal-oriented Object Importance Estimation in On-road Driving Videoscited by 19
  21. 2022Important Object Identification with Semi-Supervised Learning for Autonomous Drivingcited by 13
  22. 2015Filtered Channel Features for Pedestrian Detectioncited by 11
  23. 2021DROID: Driver-centric Risk Object Identificationcited by 9
  24. 2024Disentangled Neural Relational Inference for Interpretable Motion Predictioncited by 7
  25. 2020Where, What, Whether: Multi-modal Learning Meets Pedestrian Detectioncited by 6
  26. 2019A Self Validation Network for Object-Level Human Attention Estimationcited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026CoRE: Weakly Supervised Coarse-to-Fine Risk Evidence Learning in Driving Videoscited by 0