BiTraP: Bi-directional Pedestrian Trajectory Prediction with Multi-modal Goal Estimation
This work addresses trajectory prediction for autonomous driving and robot navigation, offering incremental improvements in accuracy for collision avoidance and navigation systems.
The paper tackled the problem of accumulated errors in long-term pedestrian trajectory prediction by proposing BiTraP, a goal-conditioned bi-directional method based on CVAE, which improved accuracy by ~10-50% over state-of-the-art results in both first-person and bird's-eye view scenarios.
Pedestrian trajectory prediction is an essential task in robotic applications such as autonomous driving and robot navigation. State-of-the-art trajectory predictors use a conditional variational autoencoder (CVAE) with recurrent neural networks (RNNs) to encode observed trajectories and decode multi-modal future trajectories. This process can suffer from accumulated errors over long prediction horizons (>=2 seconds). This paper presents BiTraP, a goal-conditioned bi-directional multi-modal trajectory prediction method based on the CVAE. BiTraP estimates the goal (end-point) of trajectories and introduces a novel bi-directional decoder to improve longer-term trajectory prediction accuracy. Extensive experiments show that BiTraP generalizes to both first-person view (FPV) and bird's-eye view (BEV) scenarios and outperforms state-of-the-art results by ~10-50%. We also show that different choices of non-parametric versus parametric target models in the CVAE directly influence the predicted multi-modal trajectory distributions. These results provide guidance on trajectory predictor design for robotic applications such as collision avoidance and navigation systems.