# Space-time Mixing Attention for Video Transformer

arXiv: https://arxiv.org/abs/2106.05968 · HTML: https://mlsanity.com/paper/2106.05968/

## Key facts

- Cited by 84 papers in the map (data through 14 Oct 2026).
- Builds on 48 of its 57 references that are in the map.
- First posted on arXiv 10 Jun 2021.

## Builds on (most cited first, 30 of 48)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Rethinking the Inception Architecture for Computer Vision](https://mlsanity.com/paper/1512.00567/) (2015)
- [mixup: Beyond Empirical Risk Minimization](https://mlsanity.com/paper/1710.09412/) (2017)
- [SGDR: Stochastic Gradient Descent with Warm Restarts](https://mlsanity.com/paper/1608.03983/) (2016)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [Longformer: The Long-Document Transformer](https://mlsanity.com/paper/2004.05150/) (2020)
- [Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset](https://mlsanity.com/paper/1705.07750/) (2017)
- [Aggregated Residual Transformations for Deep Neural Networks](https://mlsanity.com/paper/1611.05431/) (2016)
- [Non-local Neural Networks](https://mlsanity.com/paper/1711.07971/) (2017)
- [The Kinetics Human Action Video Dataset](https://mlsanity.com/paper/1705.06950/) (2017)
- [SlowFast Networks for Video Recognition](https://mlsanity.com/paper/1812.03982/) (2018)
- [Big Bird: Transformers for Longer Sequences](https://mlsanity.com/paper/2007.14062/) (2020)
- [Is Space-Time Attention All You Need for Video Understanding?](https://mlsanity.com/paper/2102.05095/) (2021)
- [ViViT: A Video Vision Transformer](https://mlsanity.com/paper/2103.15691/) (2021)
- [Generating Long Sequences with Sparse Transformers](https://mlsanity.com/paper/1904.10509/) (2019)
- [Random Erasing Data Augmentation](https://mlsanity.com/paper/1708.04896/) (2017)
- [A Closer Look at Spatiotemporal Convolutions for Action Recognition](https://mlsanity.com/paper/1711.11248/) (2017)
- [The "something something" video database for learning and evaluating visual common sense](https://mlsanity.com/paper/1706.04261/) (2017)
- [Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet](https://mlsanity.com/paper/2101.11986/) (2021)
- [Temporal Segment Networks: Towards Good Practices for Deep Action Recognition](https://mlsanity.com/paper/1608.00859/) (2016)
- [TSM: Temporal Shift Module for Efficient Video Understanding](https://mlsanity.com/paper/1811.08383/) (2018)
- [Self-Attention Generative Adversarial Networks](https://mlsanity.com/paper/1805.08318/) (2018)
- [Scaling Egocentric Vision: The EPIC-KITCHENS Dataset](https://mlsanity.com/paper/1804.02748/) (2018)
- [Image Transformer](https://mlsanity.com/paper/1802.05751/) (2018)
- [AutoAugment: Learning Augmentation Policies from Data](https://mlsanity.com/paper/1805.09501/) (2018)

## Built on by (most cited first, 30 of 84)

- [Prompting Visual-Language Models for Efficient Video Understanding](https://mlsanity.com/paper/2112.04478/) (2021)
- [UniFormer: Unifying Convolution and Self-attention for Visual Recognition](https://mlsanity.com/paper/2201.09450/) (2022)
- [ST-Adapter: Parameter-Efficient Image-to-Video Transfer Learning](https://mlsanity.com/paper/2206.13559/) (2022)
- [UniFormer: Unified Transformer for Efficient Spatiotemporal Representation Learning](https://mlsanity.com/paper/2201.04676/) (2022)
- [Adaptive Token Sampling For Efficient Vision Transformers](https://mlsanity.com/paper/2111.15667/) (2021)
- [PhysFormer: Facial Video-based Physiological Measurement with Temporal Difference Transformer](https://mlsanity.com/paper/2111.12082/) (2021)
- [EdgeViTs: Competing Light-weight CNNs on Mobile Devices with Vision Transformers](https://mlsanity.com/paper/2205.03436/) (2022)
- [UniFormerV2: Spatiotemporal Learning by Arming Image ViTs with Video UniFormer](https://mlsanity.com/paper/2211.09552/) (2022)
- [TS2-Net: Token Shift and Selection Transformer for Text-Video Retrieval](https://mlsanity.com/paper/2207.07852/) (2022)
- [Masked Video Distillation: Rethinking Masked Feature Modeling for Self-supervised Video Representation Learning](https://mlsanity.com/paper/2212.04500/) (2022)
- [Video Transformers: A Survey](https://mlsanity.com/paper/2201.05991/) (2022)
- [Revisiting Temporal Modeling for CLIP-based Image-to-Video Knowledge Transferring](https://mlsanity.com/paper/2301.11116/) (2023)
- [Open-VCLIP: Transforming CLIP to an Open-vocabulary Video Model via Interpolated Weight Optimization](https://mlsanity.com/paper/2302.00624/) (2023)
- [InternVideo-Ego4D: A Pack of Champion Solutions to Ego4D Challenges](https://mlsanity.com/paper/2211.09529/) (2022)
- [PhysFormer++: Facial Video-based Physiological Measurement with SlowFast Temporal Difference Transformer](https://mlsanity.com/paper/2302.03548/) (2023)
- [With a Little Help from my Temporal Context: Multimodal Egocentric Action Recognition](https://mlsanity.com/paper/2111.01024/) (2021)
- [Disentangling Spatial and Temporal Learning for Efficient Image-to-Video Transfer Learning](https://mlsanity.com/paper/2309.07911/) (2023)
- [Multimodal Distillation for Egocentric Action Recognition](https://mlsanity.com/paper/2307.07483/) (2023)
- [Prune Spatio-temporal Tokens by Semantic-aware Temporal Accumulation](https://mlsanity.com/paper/2308.04549/) (2023)
- [Building an Open-Vocabulary Video CLIP Model with Better Architectures, Optimization and Data](https://mlsanity.com/paper/2310.05010/) (2023)
- [OmniVid: A Generative Framework for Universal Video Understanding](https://mlsanity.com/paper/2403.17935/) (2024)
- [MorphMLP: An Efficient MLP-Like Backbone for Spatial-Temporal Representation Learning](https://mlsanity.com/paper/2111.12527/) (2021)
- [Cross-Modal Learning with 3D Deformable Attention for Action Recognition](https://mlsanity.com/paper/2212.05638/) (2022)
- [Efficient Video Action Detection with Token Dropout and Context Refinement](https://mlsanity.com/paper/2304.08451/) (2023)
- [SCOTCH and SODA: A Transformer Video Shadow Detection Framework](https://mlsanity.com/paper/2211.06885/) (2022)
- [AutoLabel: CLIP-based framework for Open-set Video Domain Adaptation](https://mlsanity.com/paper/2304.01110/) (2023)
- [COMEDIAN: Self-Supervised Learning and Knowledge Distillation for Action Spotting using Transformers](https://mlsanity.com/paper/2309.01270/) (2023)
- [EAN: Event Adaptive Network for Enhanced Action Recognition](https://mlsanity.com/paper/2107.10771/) (2021)
- [An Efficient Spatio-Temporal Pyramid Transformer for Action Detection](https://mlsanity.com/paper/2207.10448/) (2022)
- [TimeBalance: Temporally-Invariant and Temporally-Distinctive Video Representations for Semi-Supervised Action Recognition](https://mlsanity.com/paper/2303.16268/) (2023)

## Abstract (arXiv)

This paper is on video recognition using Transformers. Very recent attempts in this area have demonstrated promising results in terms of recognition accuracy, yet they have been also shown to induce, in many cases, significant computational overheads due to the additional modelling of the temporal information. In this work, we propose a Video Transformer model the complexity of which scales linearly with the number of frames in the video sequence and hence induces no overhead compared to an image-based Transformer model. To achieve this, our model makes two approximations to the full space-time attention used in Video Transformers: (a) It restricts time attention to a local temporal window and capitalizes on the Transformer's depth to obtain full temporal coverage of the video sequence. (b) It uses efficient space-time mixing to attend jointly spatial and temporal locations without inducing any additional cost on top of a spatial-only attention model. We also show how to integrate 2 very lightweight mechanisms for global temporal-only attention which provide additional accuracy improvements at minimal computational cost. We demonstrate that our model produces very high recognition accuracy on the most popular video recognition datasets while at the same time being significantly more efficient than other Video Transformer models. Code will be made available.