KING: Generating Safety-Critical Driving Scenarios
for Robust Imitation via Kinematics Gradients

Niklas Hanselmann 1,2,3 Katrin Renz 2,3 Kashyap Chitta 2,3
Apratim Bhattacharyya 2,3 Andreas Geiger 2,3
1 Mercedes-Benz AG R&D 2 University of Tübingen
3 Max Planck Institute for Intelligent Systems Tübingen
ECCV 2022 (oral)


Suppl. Material


ECCV Video

Suppl. Video

TL;DR: Previous works on the automated generation of safety-critical driving scenarios have resorted to black-box optimization (BBO) techniques to handle non-differentiable simulators and driving agents. We propose an approximation involving a differentiable kinematics model of the simulators true dynamics that enables more efficient generation via backpropagation. Using the generated scenarios as additional training data, we improve the robustness of an imitation learning-based driving agent.

Simulators offer the possibility of safe, low-cost development of self-driving systems. However, current driving simulators exhibit naïve behavior models for background traffic. Hand-tuned scenarios are typically added during simulation to induce safety-critical situations. An alternative approach is to adversarially perturb the background traffic trajectories. In this paper, we study this approach to safety-critical driving scenario generation using the CARLA simulator. We use a kinematic bicycle model as a proxy to the simulator's true dynamics and observe that gradients through this proxy model are sufficient for optimizing the background traffic trajectories. Based on this finding, we propose KING, which generates safety-critical driving scenarios with a 20% higher success rate than black-box optimization. By solving the scenarios generated by KING using a privileged rule-based expert algorithm, we obtain training data for an imitation learning policy. After fine-tuning on this new data, we show that the policy becomes better at avoiding collisions. Importantly, our generated data leads to reduced collisions on both held-out scenarios generated via KING as well as traditional hand-crafted scenarios, demonstrating improved robustness.
Qualitative Examples


This work was supported by the German Federal Ministry for Economic Affairs and Climate Action within the project KI Delta Learning (project numbers: 19A19013A, 19A19013O), the German Federal Ministry of Education and Research (Tübingen AI Center, FKZ: 01IS18039A, 01IS18039B) and the German Research Foundation (SFB 1233, Robust Vision: Inference Principles and Neural Mechanisms, TP 17, project number: 276693517). We thank the International Max Planck Research School for Intelligent Systems (IMPRS-IS) for supporting Katrin Renz and Kashyap Chitta. The authors also thank Aditya Prakash and Bernhard Jaeger for proofreading.

The template for this website was borrowed and adapted from Despoina Paschalidou.