Self-Supervised Poisson-Gaussian Denoising

Khademi, Wesley; Rao, Sonia; Minnerath, Clare; Hagen, Guy M.; Ventura, Jonathan

doi:10.1109/wacv48630.2021.00218

Cited by 20 publications

(6 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Novel strategies of self-supervision are being developed and employed for different problems such as multimodal learning [86, 62], self-labeling [56, 48], learning semantic context [19] and contrastive predictive coding [4]. Denoising strategies based on self-supervision have revolutionized image denoising performance across different scientific domains[45, 40]. Leveraging statistical independence of noise, first introduced in the work of Noise2Noise (N2N) [53] was given a theoretical grounding in the work of Noise2Self by posing it under a self-supervised framework [7].…”

Section: Introductionmentioning

confidence: 99%

Patch2Self denoising of Diffusion MRI with Self-Supervision and Matrix Sketching

Fadnavis¹,

Chowdhury²,

Batson

et al. 2022

Preprint

View full text Add to dashboard Cite

Diffusion-weighted magnetic resonance imaging (DWI) is the only noninvasive method for quantifying microstructure and reconstructing white-matter pathways in the living human brain. Fluctuations from multiple sources create significant additive noise in DWI data which must be suppressed before subsequent microstructure analysis. We introduce a self-supervised learning method for denoising DWI data, Patch2Self (P2S), which uses the entire volume to learn a full-rank locally linear denoiser for that volume. By taking advantage of the oversampled $q$-space of DWI data, P2S can separate structure from noise without requiring an explicit model for either. The setup of P2S however can be resource intensive, both in terms of running time and memory usage, as it uses all voxels (n) from all-but-one held-in volumes (d-1) to learn a linear mapping $\Phi : \mathbb{R}^{n \times (d-1)} \mapsto \mathbb{R}^{n}$ for denoising the held-out volume. We exploit the redundancy imposed by P2S to alleviate its performance issues and inspect regions that influence the noise disproportionately. Specifically we introduce P2S-sketch, which makes a two-fold contribution: \textit{(1)} P2S-sketch uses matrix sketching to perform self-supervised denoising. By solving a sub-problem on a smaller sub-space, so called, \textit{coreset}, we show how P2S can yield a significant speedup in training time while using less memory. \textit{(2)} We show how the so-called statistical leverage scores can be used to \textit{interpret} the denoising of dMRI data, a process that was traditionally treated as a black-box. Our experiments conducted on simulated and real data clearly demonstrate that P2S via matrix sketching (P2S-sketch) does not lead to any loss in denoising quality, while yielding significant speedup and improved memory usage by training on a smaller fraction of the data. With thorough comparisons on real and simulated data, we show that Patch2Self outperforms the current state-of-the-art methods for DWI denoising both in terms of visual conspicuity and downstream modeling tasks. We demonstrate the effectiveness of our approach via multiple quantitative metrics such as fiber bundle coherence, $R^2$ via cross-validation on model fitting, mean absolute error of DTI residuals across a cohort of sixty subjects.

show abstract

Section: Introductionmentioning

confidence: 99%

Patch2Self denoising of Diffusion MRI with Self-Supervision and Matrix Sketching

Fadnavis¹,

Chowdhury²,

Batson

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Supervised methods require training the network on a dataset prior to performing denoising on experimental images and include content-aware image restoration (CARE) [17], denoising convolutional neural networks (DnCNN) [18], and Noise2Noise (N2N) [19]. Self-supervised methods train the network on the same image it is performing denoising on and Noise2Void (N2V) [20], probabilistic Noise2Void (PN2V) [21], Noise2Fast [22], and other methods [23, 24]. Supervised CNN training methods typically provide higher performance than the self-supervised methods due to the inclusion of additional information in target images.…”

Section: Introductionmentioning

confidence: 99%

Real-time Image Denoising of Mixed Poisson-Gaussian Noise in Fluorescence Microscopy Images using ImageJ

Mannam

Zhang

Zhu

et al. 2021

Preprint

View full text Add to dashboard Cite

Fluorescence microscopy imaging speed is fundamentally limited by the measurement signal-to-noise ratio (SNR). To improve image SNR for a given image acquisition rate, computational denoising techniques can be used to suppress noise. However, analytical techniques to estimate a denoised image from a single frame are either computationally expensive or rely on simple noise statistical models. These models assume Poisson or Gaussian noise statistics, which are not appropriate for many fluorescence microscopy applications that contain quantum shot noise and electronic Johnson–Nyquist noise, and therefore a mixture of Poisson and Gaussian noise. In this paper, we show convolutional neural networks (CNNs) trained on mixed Poisson and Gaussian noise images to overcome the limitations of existing image denoising methods. The trained CNN is presented as an open-source ImageJ plugin that performs instant image denoising (within tens of milliseconds) with superior performance (∼8.1 dB SNR improvement) compared to the conventional fluorescence microscopy denoising methods. The method is validated on external datasets with out-of-distribution noise and contrast from the training data and consistently achieves high performance (> 8dB) denoising in less time than other fluorescence microscopy denoising methods.

show abstract

“…However, due to the weak signals and diffraction limit, FM images suffer from high noise artifacts. There are several methods focused on the development of DL-based denoising schemes in FM imaging [247][248][249]. Weigert et al [247] applied a data generation technique to collect semi-synthetic FM images followed by training a UNet model for image restoration.…”

Section: Positron Emission Tomography Positron Emission Tomography (Pet)mentioning

confidence: 99%

Image denoising in the deep learning era

2022

View full text Add to dashboard Cite

Over the last decade, the number of digital images captured per day witnessed a massive explosion. Nevertheless, the visual quality of photographs is often degraded by noise during image acquisition or transmission. With the re-emergence of deep neural networks, the performance of image denoising techniques has been substantially improved in recent years. The objective of this paper is to provide a comprehensive survey of recent advances in image denoising techniques based on deep neural networks. In doing so, we commence with a thorough description of the fundamental preliminaries of the image denoising problem followed by highlighting the benchmark datasets and the widely used metrics for objective assessments. Subsequently, we study the existing deep denoisers in the supervised and unsupervised categories and review the technical specifics of some representative methods within each category. Last but not least, we conclude the analysis by remarking on trends and challenges in the development of better state-of-the-art algorithms and future research.

show abstract

Self-Supervised Poisson-Gaussian Denoising

Cited by 20 publications

References 18 publications

Patch2Self denoising of Diffusion MRI with Self-Supervision and Matrix Sketching

Patch2Self denoising of Diffusion MRI with Self-Supervision and Matrix Sketching

Real-time Image Denoising of Mixed Poisson-Gaussian Noise in Fluorescence Microscopy Images using ImageJ

Image denoising in the deep learning era

Contact Info

Product

Resources

About