LGMASYMLMar 24, 2020

Towards Safer Self-Driving Through Great PAIN (Physically Adversarial Intelligent Networks)

arXiv:2003.10662v120 citations
Originality Incremental advance
AI Analysis

This work addresses safety issues in self-driving vehicles by generating adversarial scenarios to enhance training, though it is incremental as it builds on existing adversarial methods in simulation.

The paper tackled the problem of improving self-driving vehicle safety by addressing overfit and poor generalizability in neural networks due to limited data. It introduced a Physically Adversarial Intelligent Network (PAIN) in simulation, resulting in a protagonist agent that became more resilient to environmental uncertainty and less prone to corner case failures, with increased mean-time-to-failure and distance traveled under non-hostile conditions.

Automated vehicles' neural networks suffer from overfit, poor generalizability, and untrained edge cases due to limited data availability. Researchers synthesize randomized edge-case scenarios to assist in the training process, though simulation introduces potential for overfit to latent rules and features. Automating worst-case scenario generation could yield informative data for improving self driving. To this end, we introduce a "Physically Adversarial Intelligent Network" (PAIN), wherein self-driving vehicles interact aggressively in the CARLA simulation environment. We train two agents, a protagonist and an adversary, using dueling double deep Q networks (DDDQNs) with prioritized experience replay. The coupled networks alternately seek-to-collide and to avoid collisions such that the "defensive" avoidance algorithm increases the mean-time-to-failure and distance traveled under non-hostile operating conditions. The trained protagonist becomes more resilient to environmental uncertainty and less prone to corner case failures resulting in collisions than the agent trained without an adversary.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes