Group Activity Prediction with Sequential Relational Anticipation Model

TL;DR

This paper introduces a sequential relational anticipation model that predicts group activities from incomplete observations by modeling relational dynamics and positions, significantly outperforming existing methods.

Contribution

The paper presents a novel SRAM that explicitly anticipates activity features and positions using graph auto-encoders for improved group activity prediction.

Findings

Outperforms state-of-the-art activity prediction methods

Effectively models relational dynamics and positional information

Demonstrates robustness on multiple datasets

Abstract

In this paper, we propose a novel approach to predict group activities given the beginning frames with incomplete activity executions. Existing action prediction approaches learn to enhance the representation power of the partial observation. However, for group activity prediction, the relation evolution of people's activity and their positions over time is an important cue for predicting group activity. To this end, we propose a sequential relational anticipation model (SRAM) that summarizes the relational dynamics in the partial observation and progressively anticipates the group representations with rich discriminative information. Our model explicitly anticipates both activity features and positions by two graph auto-encoders, aiming to learn a discriminative group representation for group activity prediction. Experimental results on two popularly used datasets demonstrate that our approach significantly outperforms the state-of-the-art activity prediction methods.