Towards More Practical Group Activity Detection: A New Benchmark and Model

TL;DR

This paper introduces a new large-scale, practical dataset called Café for group activity detection and proposes a novel model that effectively handles unknown group numbers and members, outperforming previous methods.

Contribution

The paper presents a new dataset for practical GAD scenarios and a novel model that improves accuracy and speed in detecting groups and activities.

Findings

Our model outperforms previous methods in accuracy.

The model achieves faster inference speeds.

The Café dataset provides more realistic GAD benchmarks.

Abstract

Group activity detection (GAD) is the task of identifying members of each group and classifying the activity of the group at the same time in a video. While GAD has been studied recently, there is still much room for improvement in both dataset and methodology due to their limited capability to address practical GAD scenarios. To resolve these issues, we first present a new dataset, dubbed Caf\'e. Unlike existing datasets, Caf\'e is constructed primarily for GAD and presents more practical scenarios and metrics, as well as being large-scale and providing rich annotations. Along with the dataset, we propose a new GAD model that deals with an unknown number of groups and latent group members efficiently and effectively. We evaluated our model on three datasets including Caf\'e, where it outperformed previous work in terms of both accuracy and inference speed.