Ellipse Loss for Scene-Compliant Motion Prediction

TL;DR

This paper introduces an ellipse loss function that improves scene compliance and trajectory realism in motion prediction models for autonomous driving by penalizing off-road predictions and considering actor dimensions.

Contribution

The novel ellipse loss enables better reasoning about scene constraints and enhances trajectory accuracy in autonomous vehicle motion prediction models.

Findings

Improved accuracy of trajectory predictions.

Enhanced scene compliance and realism.

Better handling of actor dimensions and orientation.

Abstract

Motion prediction is a critical part of self-driving technology, responsible for inferring future behavior of traffic actors in autonomous vehicle's surroundings. In order to ensure safe and efficient operations, prediction models need to output accurate trajectories that obey the map constraints. In this paper, we address this task and propose a novel ellipse loss that allows the models to better reason about scene compliance and predict more realistic trajectories. Ellipse loss penalizes off-road predictions directly in a supervised manner, by projecting the output trajectories into the top-down map frame using a differentiable trajectory rasterizer module. Moreover, it takes into account actor dimensions and orientation, providing more direct training signals to the model. We applied ellipse loss to a recently proposed state-of-the-art joint detection-prediction model to showcase its benefits. Evaluation on large-scale autonomous driving data strongly indicates that the method allows for more accurate and more realistic trajectory predictions.