mlsanityData to 14 Oct 2026

How corner is a corner case? Percentile control for highway scenario generation

arXiv 2610.05003 ↗First posted 4 Oct 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Generating corner-case scenarios with appropriate adversity in a simulation environment is critical for testing an autonomous vehicle (AV) software stack's safety performance before deployment. Existing autonomous-driving scenario generators can enforce specific behavior, adversity, or feasibility conditions, but they provide limited control over how extreme a generated scenario is relative to plausible futures in the same traffic context. This study represents the adversity of a generated scenario as its percentile in the conditional distribution of future risk given the observed history. This view supports calibrated answers to two questions: how "corner" a generated corner-case scenario is and how its "cornerness" can be fine-tuned. To this end, we formulate history-conditioned risk-percentile requests and learn a reference risk distribution that maps each requested percentile to a physical risk target. We then use a percentile-conditioned joint diffusion model with sampling-time risk guidance to generate multi-agent futures, together with a reference-based criterion for evaluating percentile realization. Experiments use the minimum post-encroachment time (PET) between the ego and its surrounding vehicles as the risk surrogate on highD. On the primary evaluation set, our method realizes 1,422 of 1,440 requests within a 0.05 percentile tolerance (98.75%), with mean percentile error 0.00673 and PET-target error 0.00991 seconds. The resulting interface connects context-relative risk specification, physical realization, and evaluation through a common risk scale. Project website and videos of generated scenarios are available at https://hhj233.github.io/CornerPercentile/.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 30 of its 49 references that are in the map. First posted on arXiv 4 Oct 2026.

Builds on

30 of 49 references.

References · 30

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2020Denoising Diffusion Implicit Modelscited by 9,781
  3. 2022Classifier-Free Diffusion Guidancecited by 5,847
  4. 2018Accurate Uncertainties for Deep Learning Using Calibrated Regressioncited by 422
  5. 2021MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learningcited by 290
  6. 2018The highD Dataset: A Drone Dataset of Naturalistic Vehicle Trajectories on German Highways for Validation of Highly Automated Driving Systemscited by 277
  7. 2023MotionDiffuser: Controllable Multi-Agent Motion Prediction using Diffusioncited by 213
  8. 2022Guided Conditional Diffusion for Controllable Traffic Simulationcited by 190
  9. 2021Generating Useful Accident-Prone Driving Scenarios via a Learned Traffic Priorcited by 139
  10. 2021AdvSim: Generating Safety-Critical Scenarios for Self-Driving Vehiclescited by 137
  11. 2023Language-Guided Traffic Simulation via Scene-Level Diffusioncited by 111
  12. 2022TrafficGen: Learning to Generate Diverse and Realistic Traffic Scenarioscited by 106
18 more
  1. 2022A Survey on Safety-Critical Driving Scenario Generation -- A Methodological Perspectivecited by 105
  2. 2019Distribution Calibration for Regressioncited by 89
  3. 2023Language Conditioned Traffic Generationcited by 87
  4. 2018Scenic: A Language for Scenario Specification and Scene Generationcited by 83
  5. 2018Scalable End-to-End Autonomous Vehicle Testing via Rare-event Simulationcited by 77
  6. 2019Adaptive Stress Testing for Autonomous Vehiclescited by 71
  7. 2023CAT: Closed-loop Adversarial Training for Safe End-to-End Drivingcited by 54
  8. 2018Scenarios for Development, Test and Validation of Automated Vehiclescited by 48
  9. 2016Accelerated Evaluation of Automated Vehicles Safety in Lane Change Scenarios Based on Importance Sampling Techniquescited by 45
  10. 2023SAFE-SIM: Safety-Critical Closed-Loop Traffic Simulation with Diffusion-Controllable Adversariescited by 38
  11. 2020Neural Bridge Sampling for Evaluating Safety-Critical Autonomous Systemscited by 35
  12. 2021Criticality Metrics for Automated Driving: A Review and Suitability Analysis of the State of the Artcited by 31
  13. 2021Diagnostics for Conditional Density Models and Bayesian Inference Algorithmscited by 28
  14. 2023A Large-Scale Study of Probabilistic Calibration in Neural Network Regressioncited by 27
  15. 2019Testing Scenario Library Generation for Connected and Automated Vehicles, Part I: Methodologycited by 25
  16. 2022Towards Instance-Wise Calibration: Local Amortized Diagnostics and Reshaping of Conditional Densities (LADaR)cited by 10
  17. 2025Steerable Adversarial Scenario Generation through Test-Time Preference Alignmentcited by 9
  18. 2025RADE: Learning Risk-Adjustable Driving Environment via Multi-Agent Conditional Diffusioncited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2021Corner Case Generation and Analysis for Safety Assessment of Autonomous Vehicles
  2. 2026ScenePilot: Controllable Boundary-Driven Critical Scenario Generation for Autonomous Driving
  3. 2024Realistic Corner Case Generation for Autonomous Vehicles with Multimodal Large Language Model
  4. 2025Controllable Latent Diffusion for Traffic Simulation
  5. 2023Scenario Diffusion: Controllable Driving Scenario Generation With Diffusion
  6. 2026Safety-Critical Scenarios Emerge from Initial Scenes ↗
  7. 2026Human Behavior-Informed Crash Scenario Generation with Real-World Crash Priors for Autonomous Vehicle Safety Evaluation
  8. 2024A novel framework for adaptive stress testing of autonomous vehicles in multi-lane roads