Video Swin Transformer
arXiv:2106.13230
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.
References in corpus (5)
Cited by in corpus (21)
- Florence: A New Foundation Model for Computer Vision
- Swin Transformer V2: Scaling Up Capacity and Resolution
- VIOLET : End-to-End Video-Language Transformers with Masked Visual-token Modeling
- SimMIM: A Simple Framework for Masked Image Modeling
- Revisiting 3D ResNets for Video Recognition
- AdaViT: Adaptive Vision Transformers for Efficient Image Recognition
- EfficientPhys: Enabling Simple, Fast and Accurate Camera-Based Vitals Measurement
- PhysFormer: Facial Video-based Physiological Measurement with Temporal Difference Transformer
- TokenLearner: What Can 8 Learned Tokens Do for Images and Videos?
- Video-Text Pre-training with Learned Regions
- VidTr: Video Transformer Without Convolutions
- Evaluating Transformers for Lightweight Action Recognition
- PatchFormer: An Efficient Point Transformer with Patch Attention
- Efficient Video Transformers with Spatial-Temporal Token Selection
- StarVQA: Space-Time Attention for Video Quality Assessment
- Semantic Segmentation on VSPW Dataset through Aggregation of Transformer Models
- Learning Higher-Order Dynamics in Video-Based Cardiac Measurement
- Overview of Tencent Multi-modal Ads Video Understanding Challenge
- "Knights": First Place Submission for VIPriors21 Action Recognition Challenge at ICCV 2021
- Technical Report: Disentangled Action Parsing Networks for Accurate Part-level Action Parsing
- Ripple Attention for Visual Perception with Sub-quadratic Complexity