Video Swin Transformer

TL;DR

This paper introduces a locality-based Video Swin Transformer architecture that improves speed and accuracy in video recognition tasks by adapting the Swin Transformer for video data, leveraging local attention instead of global self-attention.

Contribution

The paper proposes a novel Video Swin Transformer that incorporates locality bias, achieving state-of-the-art results with better efficiency compared to global attention models.

Findings

Achieves 84.9% top-1 accuracy on Kinetics-400

Uses ~20x less pre-training data and smaller models

Attains 69.6% top-1 accuracy on Something-Something v2

Abstract

The vision community is witnessing a modeling shift from CNNs to Transformers, where pure Transformer architectures have attained top accuracy on the major video recognition benchmarks. These video models are all built on Transformer layers that globally connect patches across the spatial and temporal dimensions. In this paper, we instead advocate an inductive bias of locality in video Transformers, which leads to a better speed-accuracy trade-off compared to previous approaches which compute self-attention globally even with spatial-temporal factorization. The locality of the proposed video architecture is realized by adapting the Swin Transformer designed for the image domain, while continuing to leverage the power of pre-trained image models. Our approach achieves state-of-the-art accuracy on a broad range of video recognition benchmarks, including on action recognition (84.9 top-1 accuracy on Kinetics-400 and 86.1 top-1 accuracy on Kinetics-600 with ~20x less pre-training data and ~3x smaller model size) and temporal modeling (69.6 top-1 accuracy on Something-Something v2). The code and models will be made publicly available at https://github.com/SwinTransformer/Video-Swin-Transformer.