TrajRS: Towards Certified Robustness in Pedestrian Trajectory Prediction
This work addresses the need for verifiable safety assurances in autonomous driving by providing certified robustness for trajectory prediction models.
The paper introduces TrajRS, a certified robustness framework for pedestrian trajectory prediction that extends Randomized Smoothing to provide a certified robust radius. Experiments show TrajRS effectively certifies robustness for all smoothed predictors.
The robustness of trajectory prediction models is crucial for developing safe autonomous driving systems. Adversarial attacks on trajectory prediction can significantly impair the accuracy of predicted trajectories, leading to hazardous driving behaviors. While heuristic defense strategies have been implemented to enhance the robustness of trajectory prediction models, these measures often fail against more sophisticated, targeted adversarial attacks. Hence, there is a pressing need to establish verifiable safety assurances for trajectory prediction models. In this paper, we extend the traditional Randomized Smoothing framework to "TrajRS", which provides a certified robust radius for smoothed trajectory predictors. We clarify and expand the formal definitions of robustness in trajectory prediction and tailor the practical TrajRS scheme specifically to "robustness for the optimal prediction" and "robustness for all possible predictions". An extensive set of experiments demonstrates that TrajRS effectively achieves robustness certification for all smoothed pedestrian trajectory predictors in this work.