Improving Out-of-Distribution Generalization of Trajectory Prediction for Autonomous Driving via Polynomial Representations
This addresses robustness issues in autonomous driving trajectory prediction, though it is incremental as it builds on existing models with a new representation and testing protocol.
The paper tackles the problem of improving out-of-distribution generalization for trajectory prediction in autonomous driving by introducing a polynomial representation-based algorithm, achieving near state-of-the-art in-distribution performance and significantly better robustness with reduced model size, training effort, and inference time.
Robustness against Out-of-Distribution (OoD) samples is a key performance indicator of a trajectory prediction model. However, the development and ranking of state-of-the-art (SotA) models are driven by their In-Distribution (ID) performance on individual competition datasets. We present an OoD testing protocol that homogenizes datasets and prediction tasks across two large-scale motion datasets. We introduce a novel prediction algorithm based on polynomial representations for agent trajectory and road geometry on both the input and output sides of the model. With a much smaller model size, training effort, and inference time, we reach near SotA performance for ID testing and significantly improve robustness in OoD testing. Within our OoD testing protocol, we further study two augmentation strategies of SotA models and their effects on model generalization. Highlighting the contrast between ID and OoD performance, we suggest adding OoD testing to the evaluation criteria of trajectory prediction models.