Learning Enriched Features for Real Image Restoration and Enhancement

Authors

Syed Waqas Zamir, Inception Institute of Artificial Intelligence
Aditya Arora, Inception Institute of Artificial IntelligenceFollow
Salman Khan, Inception Institute of Artificial Intelligence & Mohamed bin Zayed University of Artificial IntelligenceFollow
Munawar Hayat, Inception Institute of Artificial Intelligence & Mohamed bin Zayed University of Artificial Intelligence
Fahad Shabaz Khan, Inception Institute of Artificial Intelligence & Mohamed bin Zayed University of Artificial IntelligenceFollow
Ming-Hsuan Yang, University of California, Merced, United States & Google Research, Merced, United States
Ling Shao, Inception Institute of Artificial Intelligence & Mohamed bin Zayed University of Artificial IntelligenceFollow

Document Type

Conference Proceeding

Publication Title

Lecture Notes in Computer Science

Abstract

With the goal of recovering high-quality image content from its degraded version, image restoration enjoys numerous applications, such as in surveillance, computational photography and medical imaging. Recently, convolutional neural networks (CNNs) have achieved dramatic improvements over conventional approaches for image restoration task. Existing CNN-based methods typically operate either on full-resolution or on progressively low-resolution representations. In the former case, spatially precise but contextually less robust results are achieved, while in the latter case, semantically reliable but spatially less accurate outputs are generated. In this paper, we present an architecture with the collective goals of maintaining spatially-precise high-resolution representations through the entire network and receiving strong contextual information from the low-resolution representations. The core of our approach is a multi-scale residual block containing several key elements: (a) parallel multi-resolution convolution streams for extracting multi-scale features, (b) information exchange across the multi-resolution streams, (c) spatial and channel attention mechanisms for capturing contextual information, and (d) attention based multi-scale feature aggregation. In a nutshell, our approach learns an enriched set of features that combines contextual information from multiple scales, while simultaneously preserving the high-resolution spatial details. Extensive experiments on five real image benchmark datasets demonstrate that our method, named as MIRNet, achieves state-of-the-art results for image denoising, super-resolution, and image enhancement. The source code and pre-trained models are available at https://github.com/swz30/MIRNet. © 2020, Springer Nature Switzerland AG.

First Page

492

Last Page

511

DOI

10.1007/978-3-030-58595-2_30

Publication Date

10-20-2020

Keywords

Color photography, Computer vision, Convolution, Convolutional neural networks, Image denoising, Image enhancement, Medical imaging, Restoration, Attention mechanisms, Benchmark datasets, Computational photography, Contextual information, Conventional approach, High quality images, Information exchanges, Multi-scale features, Image reconstruction

Comments

IR Deposit conditions: non-described

Recommended Citation

S.W. Zamir, et al, "Learning Enriched Features for Real Image Restoration and Enhancement", in Computer Vision – ECCV 2020, in Lecture Notes in Computer Science, vol 12370, Oct. 2020, pp. 492-511, doi.org:10.1007/978-3-030-58595-2_30