# Space-Time Attention with Shifted Non-Local Search

arXiv: https://arxiv.org/abs/2309.16849 · HTML: https://mlsanity.com/paper/2309.16849/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 34 references that are in the map.
- First posted on arXiv 28 Sep 2023.

## Builds on (most cited first, 21 of 21)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [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)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [SGDR: Stochastic Gradient Descent with Warm Restarts](https://mlsanity.com/paper/1608.03983/) (2016)
- [Squeeze-and-Excitation Networks](https://mlsanity.com/paper/1709.01507/) (2017)
- [SwinIR: Image Restoration Using Swin Transformer](https://mlsanity.com/paper/2108.10257/) (2021)
- [Deformable Convolutional Networks](https://mlsanity.com/paper/1703.06211/) (2017)
- [The 2017 DAVIS Challenge on Video Object Segmentation](https://mlsanity.com/paper/1704.00675/) (2017)
- [Optical Flow Estimation using a Spatial Pyramid Network](https://mlsanity.com/paper/1611.00850/) (2016)
- [Vision Transformer with Deformable Attention](https://mlsanity.com/paper/2201.00520/) (2022)
- [Neighborhood Attention Transformer](https://mlsanity.com/paper/2204.07143/) (2022)
- [TDAN: Temporally Deformable Alignment Network for Video Super-Resolution](https://mlsanity.com/paper/1812.02898/) (2018)
- [VRT: A Video Restoration Transformer](https://mlsanity.com/paper/2201.12288/) (2022)
- [Recurrent Video Restoration Transformer with Guided Deformable Attention](https://mlsanity.com/paper/2206.02146/) (2022)
- [FastDVDnet: Towards Real-Time Deep Video Denoising Without Flow Estimation](https://mlsanity.com/paper/1907.01361/) (2019)
- [Neural Nearest Neighbors Networks](https://mlsanity.com/paper/1810.12575/) (2018)
- [Patch Craft: Video Denoising by Deep Modeling and Patch Matching](https://mlsanity.com/paper/2103.13767/) (2021)
- [Active Convolution: Learning the Shape of Convolution for Image Classification](https://mlsanity.com/paper/1703.09076/) (2017)
- [COLA-Net: Collaborative Attention Network for Image Restoration](https://mlsanity.com/paper/2103.05961/) (2021)

## Abstract (arXiv)

Efficiently computing attention maps for videos is challenging due to the motion of objects between frames. While a standard non-local search is high-quality for a window surrounding each query point, the window's small size cannot accommodate motion. Methods for long-range motion use an auxiliary network to predict the most similar key coordinates as offsets from each query location. However, accurately predicting this flow field of offsets remains challenging, even for large-scale networks. Small spatial inaccuracies significantly impact the attention module's quality. This paper proposes a search strategy that combines the quality of a non-local search with the range of predicted offsets. The method, named Shifted Non-Local Search, executes a small grid search surrounding the predicted offsets to correct small spatial errors. Our method's in-place computation consumes 10 times less memory and is over 3 times faster than previous work. Experimentally, correcting the small spatial errors improves the video frame alignment quality by over 3 dB PSNR. Our search upgrades existing space-time attention modules, which improves video denoising results by 0.30 dB PSNR for a 7.5% increase in overall runtime. We integrate our space-time attention module into a UNet-like architecture to achieve state-of-the-art results on video denoising.