Join us for the third dScience Brain Talk on December 12 with guest speaker Aizaz Sharif, Ph.D. Researcher at Simula Research Laboratory (VIAS - Department of Validation Intelligence for Autonomous Software Systems)and University of Oslo (UiO).
The Brain Talk webinar is an online platform that gives the opportunity to scientists, researchers and early-stage researchers to present, discuss and share their brilliant ideas on Machine Learning (ML) and Computational Science. We believe that everyone should have the opportunity to learn and achieve their full potential. To that goal, innovative ideas are shared here.
Program
Testing and Improving the Robustness of Self-driving cars using Adversarial Deep Reinforcement Learning
Aizaz Sharif's Ph.D. research is focused on testing and validating autonomous driving systems that are operating using artificial intelligence-based software. He is currently working under the supervision of Dusica Marijan.
Autonomous cars are well known for being vulnerable to adversarial attacks that can compromise the safety of the car and pose danger to other road users. To effectively defend against adversaries, it is required to not only test autonomous cars for finding driving errors but to improve the robustness of the cars to these errors. To this end, in this paper, we propose a two-step methodology for autonomous cars that consists of (i) finding failure states in autonomous cars by training the adversarial driving agent, and (ii) improving the robustness of autonomous cars by retraining them with effective adversarial inputs. Our methodology supports testing autonomous cars in a multi-agent environment, where we train and compare adversarial car policy on two custom reward functions to test the driving control decision of autonomous cars. We run experiments in a vision-based high-fidelity urban driving simulated environment. Our results show that adversarial testing can be used for finding erroneous autonomous driving behavior, followed by adversarial training for improving the robustness of deep reinforcement learning-based autonomous driving policies. We demonstrate that the autonomous cars retrained using effective adversarial inputs noticeably increase the performance of their driving policies in terms of reduced collision and offroad steering errors.
This webinar series for dScience is produced and organized by The Brain Talk team: