TrajFlow: Multi-modal Motion Prediction via Flow Matching

TL;DR

TrajFlow is a flow matching-based motion prediction framework that efficiently generates multiple plausible trajectories in a single pass, improving accuracy, uncertainty estimation, and inference speed for autonomous driving.

Contribution

The paper introduces TrajFlow, a novel multi-modal motion prediction method that reduces computational cost and enhances uncertainty estimation using flow matching and self-conditioning techniques.

Findings

Achieves state-of-the-art performance on Waymo dataset

Predicts multiple trajectories in a single inference pass

Improves uncertainty estimation with ranking loss

Abstract

Efficient and accurate motion prediction is crucial for ensuring safety and informed decision-making in autonomous driving, particularly under dynamic real-world conditions that necessitate multi-modal forecasts. We introduce TrajFlow, a novel flow matching-based motion prediction framework that addresses the scalability and efficiency challenges of existing generative trajectory prediction methods. Unlike conventional generative approaches that employ i.i.d. sampling and require multiple inference passes to capture diverse outcomes, TrajFlow predicts multiple plausible future trajectories in a single pass, significantly reducing computational overhead while maintaining coherence across predictions. Moreover, we propose a ranking loss based on the Plackett-Luce distribution to improve uncertainty estimation of predicted trajectories. Additionally, we design a self-conditioning training technique that reuses the model's own predictions to construct noisy inputs during a second forward pass, thereby improving generalization and accelerating inference. Extensive experiments on the large-scale Waymo Open Motion Dataset (WOMD) demonstrate that TrajFlow achieves state-of-the-art performance across various key metrics, underscoring its effectiveness for safety-critical autonomous driving applications. The code and other details are available on the project website https://traj-flow.github.io/.