Projects per year
Abstract
With its significant performance improvements, the deep learning paradigm has become a standard tool for modern image denoisers. While promising performance has been shown on seen noise distributions, existing approaches often suffer from generalisation to unseen noise types or general and real noise. It is understandable as the model is designed to learn paired mapping (e.g. from a noisy image to its clean version). In this paper, we instead propose to learn to disentangle the noisy image, under the intuitive assumption that different corrupted versions of the same clean image share a common latent space. A self-supervised learning framework is proposed to achieve the goal, without looking at the latent clean image. By taking two different corrupted versions of the same image as input, the proposed Multi-view Self-supervised Disentanglement (MeD) approach learns to disentangle the latent clean features from the corruptions and recover the clean image consequently. Extensive experimental analysis on both synthetic and real noise shows the superiority of the proposed method over prior self-supervised approaches, especially on unseen novel noise types. On real noise, the proposed method even outperforms its supervised counterparts by over 3 dB.
Original language | English |
---|---|
Title of host publication | 2023 IEEE/CVF International Conference on Computer Vision (ICCV) |
Publisher | IEEE |
Pages | 12247-12257 |
Number of pages | 11 |
ISBN (Electronic) | 9798350307184 |
ISBN (Print) | 9798350307191 |
DOIs | |
Publication status | Published - 15 Jan 2024 |
Event | 2023 IEEE/CVF International Conference on Computer Vision (ICCV) - Paris, France Duration: 1 Oct 2023 → 6 Oct 2023 |
Publication series
Name | International Conference on Computer Vision (ICCV) |
---|---|
Publisher | IEEE |
ISSN (Print) | 1550-5499 |
ISSN (Electronic) | 2380-7504 |
Conference
Conference | 2023 IEEE/CVF International Conference on Computer Vision (ICCV) |
---|---|
Period | 1/10/23 → 6/10/23 |
Bibliographical note
Funding Information:The computations described in this research were performed using the Baskerville Tier 2 HPC service1. Baskerville was funded by the EPSRC and UKRI through the World Class Labs scheme (EP/T022221/1) and the Digital Research Infrastructure programme (EP/W032244/1) and is operated by Advanced Research Computing at the University of Birmingham.
Publisher Copyright:
© 2023 IEEE.
Keywords
- Training
- Adaptation models
- Computational modeling
- Noise reduction
- Self-supervised learning
- Performance gain
- Noise measurement
ASJC Scopus subject areas
- Software
- Computer Vision and Pattern Recognition
Fingerprint
Dive into the research topics of 'Multi-view Self-supervised Disentanglement for General Image Denoising'. Together they form a unique fingerprint.-
Baskerville: a national accelerated compute resource
Engineering & Physical Science Research Council, Lenovo UK Limited
13/10/20 → 31/03/25
Project: Research Councils
-
Baskerville 2.0: Enhanced Provision for High End and On-Demand Users
Engineering & Physical Science Research Council
4/01/22 → 3/05/22
Project: Research Councils