Rethinking Trajectory Prediction via "Team Game"

TL;DR

This paper introduces a hierarchical approach to multi-agent trajectory prediction that explicitly models group and individual interactions, significantly enhancing prediction accuracy in team sports and pedestrian scenarios.

Contribution

It proposes a novel hierarchical latent space framework that explicitly captures multi-level interactions, improving over existing implicit modeling methods.

Findings

Outperforms existing methods on team sports trajectory prediction.

Achieves superior accuracy in pedestrian multi-agent scenarios.

Effectively models complex multi-level interactions.

Abstract

To accurately predict trajectories in multi-agent settings, e.g. team games, it is important to effectively model the interactions among agents. Whereas a number of methods have been developed for this purpose, existing methods implicitly model these interactions as part of the deep net architecture. However, in the real world, interactions often exist at multiple levels, e.g. individuals may form groups, where interactions among groups and those among the individuals in the same group often follow significantly different patterns. In this paper, we present a novel formulation for multi-agent trajectory prediction, which explicitly introduces the concept of interactive group consensus via an interactive hierarchical latent space. This formulation allows group-level and individual-level interactions to be captured jointly, thus substantially improving the capability of modeling complex dynamics. On two multi-agent settings, i.e. team sports and pedestrians, the proposed framework consistently achieves superior performance compared to existing methods.