SocialVAE: Human Trajectory Prediction using Timewise Latents
This work addresses the problem of safe and efficient human-agent interactions for robotics and autonomous systems, though it appears incremental as it builds on existing VAE and attention methods.
The paper tackles the challenge of capturing uncertainty and multimodality in pedestrian trajectory prediction by proposing SocialVAE, a timewise variational autoencoder with social attention and backward posterior approximation, which improves state-of-the-art performance on benchmarks like ETH/UCY, Stanford Drone, and SportVU NBA.
Predicting pedestrian movement is critical for human behavior analysis and also for safe and efficient human-agent interactions. However, despite significant advancements, it is still challenging for existing approaches to capture the uncertainty and multimodality of human navigation decision making. In this paper, we propose SocialVAE, a novel approach for human trajectory prediction. The core of SocialVAE is a timewise variational autoencoder architecture that exploits stochastic recurrent neural networks to perform prediction, combined with a social attention mechanism and a backward posterior approximation to allow for better extraction of pedestrian navigation strategies. We show that SocialVAE improves current state-of-the-art performance on several pedestrian trajectory prediction benchmarks, including the ETH/UCY benchmark, Stanford Drone Dataset, and SportVU NBA movement dataset. Code is available at: https://github.com/xupei0610/SocialVAE.