Context-free Self-Conditioned GAN for Trajectory Forecasting

TL;DR

This paper introduces a novel context-free, self-conditioned GAN approach for trajectory forecasting that learns multiple behavioral modes without supervision, outperforming previous methods on human motion and road agent datasets.

Contribution

The paper proposes a new unsupervised, context-free GAN framework with self-conditioning for trajectory prediction, improving mode learning and forecasting accuracy.

Findings

Outperforms previous context-free methods in supervised label settings.

Achieves superior results in human motion trajectory forecasting.

Performs competitively in road agent trajectory prediction.

Abstract

In this paper, we present a context-free unsupervised approach based on a self-conditioned GAN to learn different modes from 2D trajectories. Our intuition is that each mode indicates a different behavioral moving pattern in the discriminator's feature space. We apply this approach to the problem of trajectory forecasting. We present three different training settings based on self-conditioned GAN, which produce better forecasters. We test our method in two data sets: human motion and road agents. Experimental results show that our approach outperforms previous context-free methods in the least representative supervised labels while performing well in the remaining labels. In addition, our approach outperforms globally in human motion, while performing well in road agents.