Motion Prediction using Trajectory Sets and Self-Driving Domain Knowledge

TL;DR

This paper enhances trajectory classification methods for vehicle motion prediction by incorporating map-based auxiliary losses and spatial-temporal weighting, leading to improved accuracy especially on smaller datasets.

Contribution

It introduces an auxiliary loss penalizing off-road predictions and explores weighted losses to better model spatial-temporal relationships, with a comprehensive comparison of classification and ordinal regression.

Findings

Auxiliary loss improves small dataset performance

Weighted losses better capture spatial-temporal relationships

Classification-based methods achieve state-of-the-art results

Abstract

Predicting the future motion of vehicles has been studied using various techniques, including stochastic policies, generative models, and regression. Recent work has shown that classification over a trajectory set, which approximates possible motions, achieves state-of-the-art performance and avoids issues like mode collapse. However, map information and the physical relationships between nearby trajectories is not fully exploited in this formulation. We build on classification-based approaches to motion prediction by adding an auxiliary loss that penalizes off-road predictions. This auxiliary loss can easily be pretrained using only map information (e.g., off-road area), which significantly improves performance on small datasets. We also investigate weighted cross-entropy losses to capture spatial-temporal relationships among trajectories. Our final contribution is a detailed comparison of classification and ordinal regression on two public self-driving datasets.