SPARTAN: Self-supervised Spatiotemporal Transformers Approach to Group Activity Recognition
This addresses the problem of recognizing group activities in videos for computer vision applications, offering a novel method to reduce reliance on labeled data, though it is incremental in applying transformers to this specific task.
The paper tackles group activity recognition in videos by proposing SPARTAN, a self-supervised spatiotemporal transformer approach that uses unlabeled data to match features across varying views, achieving state-of-the-art results on NBA and Volleyball datasets with significant improvements in MCA and MPCA metrics.
In this paper, we propose a new, simple, and effective Self-supervised Spatio-temporal Transformers (SPARTAN) approach to Group Activity Recognition (GAR) using unlabeled video data. Given a video, we create local and global Spatio-temporal views with varying spatial patch sizes and frame rates. The proposed self-supervised objective aims to match the features of these contrasting views representing the same video to be consistent with the variations in spatiotemporal domains. To the best of our knowledge, the proposed mechanism is one of the first works to alleviate the weakly supervised setting of GAR using the encoders in video transformers. Furthermore, using the advantage of transformer models, our proposed approach supports long-term relationship modeling along spatio-temporal dimensions. The proposed SPARTAN approach performs well on two group activity recognition benchmarks, including NBA and Volleyball datasets, by surpassing the state-of-the-art results by a significant margin in terms of MCA and MPCA metrics.