# Unsupervised Domain-Specific Deblurring using Scale-Specific Attention

arXiv: https://arxiv.org/abs/2112.06175 · HTML: https://mlsanity.com/paper/2112.06175/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 53 references that are in the map.
- First posted on arXiv 12 Dec 2021.

## Builds on (most cited first, 25 of 25)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Squeeze-and-Excitation Networks](https://mlsanity.com/paper/1709.01507/) (2017)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [Analyzing and Improving the Image Quality of StyleGAN](https://mlsanity.com/paper/1912.04958/) (2019)
- [CBAM: Convolutional Block Attention Module](https://mlsanity.com/paper/1807.06521/) (2018)
- [Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks](https://mlsanity.com/paper/1511.06434/) (2015)
- [Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring](https://mlsanity.com/paper/1612.02177/) (2016)
- [MaskGAN: Towards Diverse and Interactive Facial Image Manipulation](https://mlsanity.com/paper/1907.11922/) (2019)
- [DeblurGAN: Blind Motion Deblurring Using Conditional Adversarial Networks](https://mlsanity.com/paper/1711.07064/) (2017)
- [DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better](https://mlsanity.com/paper/1908.03826/) (2019)
- [SinGAN: Learning a Generative Model from a Single Natural Image](https://mlsanity.com/paper/1905.01164/) (2019)
- [SCA-CNN: Spatial and Channel-wise Attention in Convolutional Networks for Image Captioning](https://mlsanity.com/paper/1611.05594/) (2016)
- [Deep Stacked Hierarchical Multi-patch Network for Image Deblurring](https://mlsanity.com/paper/1904.03468/) (2019)
- [Learning a Convolutional Neural Network for Non-uniform Motion Blur Removal](https://mlsanity.com/paper/1503.00593/) (2015)
- [Spatially-Attentive Patch-Hierarchical Network for Adaptive Motion Deblurring](https://mlsanity.com/paper/2004.05343/) (2020)
- [Deep Semantic Face Deblurring](https://mlsanity.com/paper/1803.03345/) (2018)
- [A Neural Approach to Blind Motion Deblurring](https://mlsanity.com/paper/1603.04771/) (2016)
- [Rendering Natural Camera Bokeh Effect with Deep Learning](https://mlsanity.com/paper/2006.05698/) (2020)
- [Bringing Alive Blurred Moments](https://mlsanity.com/paper/1804.02913/) (2018)
- [Unsupervised Domain-Specific Deblurring via Disentangled Representations](https://mlsanity.com/paper/1903.01594/) (2019)
- [An Experimental Evaluation of Covariates Effects on Unconstrained Face Verification](https://mlsanity.com/paper/1808.05508/) (2018)
- [Deep Portrait Image Completion and Extrapolation](https://mlsanity.com/paper/1808.07757/) (2018)
- [Deep Face Deblurring](https://mlsanity.com/paper/1704.08772/) (2017)
- [Effects of Image Degradations to CNN-based Image Classification](https://mlsanity.com/paper/1810.05552/) (2018)

## Abstract (arXiv)

In the literature, coarse-to-fine or scale-recurrent approach i.e. progressively restoring a clean image from its low-resolution versions has been successfully employed for single image deblurring. However, a major disadvantage of existing methods is the need for paired data; i.e. sharpblur image pairs of the same scene, which is a complicated and cumbersome acquisition procedure. Additionally, due to strong supervision on loss functions, pre-trained models of such networks are strongly biased towards the blur experienced during training and tend to give sub-optimal performance when confronted by new blur kernels during inference time. To address the above issues, we propose unsupervised domain-specific deblurring using a scale-adaptive attention module (SAAM). Our network does not require supervised pairs for training, and the deblurring mechanism is primarily guided by adversarial loss, thus making our network suitable for a distribution of blur functions. Given a blurred input image, different resolutions of the same image are used in our model during training and SAAM allows for effective flow of information across the resolutions. For network training at a specific scale, SAAM attends to lower scale features as a function of the current scale. Different ablation studies show that our coarse-to-fine mechanism outperforms end-to-end unsupervised models and SAAM is able to attend better compared to attention models used in literature. Qualitative and quantitative comparisons (on no-reference metrics) show that our method outperforms prior unsupervised methods.