SOURCE-LINKED INTELLIGENCE

Collision Snapshot Guided Time-Reversed Safety-Critical Scenario Generation

arXiv · AI, language, vision and robotics · article · Sep 6, 2026 · UTC

The generation of safety-critical traffic scenarios is essential for training and evaluating autonomous vehicles. Prior approaches typically perturb the trajectories of existing agents in a traffic scenario using simplified adversarial objectives to induce safety-critical interactions, which can limit the plausibility and diversity of the generated scenarios. Although inserting new adversarial vehicles can alleviate this limitation, determining when and where to introduce them in a scenario-specific manner remains challenging. In this work, we introduce \underline{CO}llision \underline{S}napsh

Read original source ↗ Open in workspace

recordType

paper

region

Global