Multi-scale Temporal Fusion Transformer for Incomplete Vehicle Trajectory Prediction

TL;DR

This paper introduces the Multi-scale Temporal Fusion Transformer (MTFT), a novel framework that effectively predicts incomplete vehicle trajectories by capturing multi-scale motion features and incorporating motion continuity, outperforming existing models in real traffic scenarios.

Contribution

The paper proposes the MTFT framework with multi-scale attention and fusion modules to improve trajectory prediction under missing data conditions, a novel approach in autonomous driving.

Findings

Achieves over 39% performance improvement on the HighD dataset.

Effectively handles missing data caused by occlusion and perception failures.

Outperforms state-of-the-art models in incomplete trajectory prediction.

Abstract

Motion prediction plays an essential role in autonomous driving systems, enabling autonomous vehicles to achieve more accurate local-path planning and driving decisions based on predictions of the surrounding vehicles. However, existing methods neglect the potential missing values caused by object occlusion, perception failures, etc., which inevitably degrades the trajectory prediction performance in real traffic scenarios. To address this limitation, we propose a novel end-to-end framework for incomplete vehicle trajectory prediction, named Multi-scale Temporal Fusion Transformer (MTFT), which consists of the Multi-scale Attention Head (MAH) and the Continuity Representation-guided Multi-scale Fusion (CRMF) module. Specifically, the MAH leverages the multi-head attention mechanism to parallelly capture multi-scale motion representation of trajectory from different temporal granularities, thus mitigating the adverse effect of missing values on prediction. Furthermore, the multi-scale motion representation is input into the CRMF module for multi-scale fusion to obtain the robust temporal feature of the vehicle. During the fusion process, the continuity representation of vehicle motion is first extracted across time steps to guide the fusion, ensuring that the resulting temporal feature incorporates both detailed information and the overall trend of vehicle motion, which facilitates the accurate decoding of future trajectory that is consistent with the vehicle's motion trend. We evaluate the proposed model on four datasets derived from highway and urban traffic scenarios. The experimental results demonstrate its superior performance in the incomplete vehicle trajectory prediction task compared with state-of-the-art models, e.g., a comprehensive performance improvement of more than 39% on the HighD dataset.