Mar 15, 2022 · We show that SocialVAE improves current state-of-the-art performance on several pedestrian trajectory prediction benchmarks, including the ETH/ ...

In this paper, we propose SocialVAE, a novel approach for human trajectory prediction. The core of SocialVAE is a timewise variational autoencoder architecture ...

In this work, we focus on the task of human trajectory prediction from short- term historical observations. Traditional works predict pedestrian trajectories.

Oct 28, 2022 · It uses an attention-based mechanism to extract human navigation strategies from the social features exhibited in short-term observations, and ...

SocialVAE performs prediction for each agent independently, based on social features extracted from local observations, and can run with scenes having an ...

The core of SocialVAE is a timewise variational autoencoder architecture that exploits stochastic recurrent neural networks to perform prediction, ...

Negative Log Likelihood (NLL), the negative logarithm of the value of the predictive PDF at GT trajectories. The predictive distribution is obtained.

Duration: 5:00
Posted: Nov 1, 2022

Predicting pedestrian movement is critical for human behavior analysis and also for safe and efficient human-agent interactions.

Bayer, J., Osendorfer, C.: Learning stochastic recurrent networks. · Becker, S., Hug, R., Hübner, W., Arens, M.: An evaluation of trajectory prediction ...