Noise2Self: Blind Denoising by Self-Supervision

Joshua Batson, Loic Royer
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:524-533, 2019.

Abstract

We propose a general framework for denoising high-dimensional measurements which requires no prior on the signal, no estimate of the noise, and no clean training data. The only assumption is that the noise exhibits statistical independence across different dimensions of the measurement, while the true signal exhibits some correlation. For a broad class of functions (“$\mathcal{J}$-invariant”), it is then possible to estimate the performance of a denoiser from noisy data alone. This allows us to calibrate $\mathcal{J}$-invariant versions of any parameterised denoising algorithm, from the single hyperparameter of a median filter to the millions of weights of a deep neural network. We demonstrate this on natural image and microscopy data, where we exploit noise independence between pixels, and on single-cell gene expression data, where we exploit independence between detections of individual molecules. This framework generalizes recent work on training neural nets from noisy images and on cross-validation for matrix factorization.

Cite this Paper

BibTeX
@InProceedings{pmlr-v97-batson19a, title = {{N}oise2{S}elf: Blind Denoising by Self-Supervision}, author = {Batson, Joshua and Royer, Loic}, booktitle = {Proceedings of the 36th International Conference on Machine Learning}, pages = {524--533}, year = {2019}, editor = {Chaudhuri, Kamalika and Salakhutdinov, Ruslan}, volume = {97}, series = {Proceedings of Machine Learning Research}, month = {09--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v97/batson19a/batson19a.pdf}, url = {https://proceedings.mlr.press/v97/batson19a.html}, abstract = {We propose a general framework for denoising high-dimensional measurements which requires no prior on the signal, no estimate of the noise, and no clean training data. The only assumption is that the noise exhibits statistical independence across different dimensions of the measurement, while the true signal exhibits some correlation. For a broad class of functions (“$\mathcal{J}$-invariant”), it is then possible to estimate the performance of a denoiser from noisy data alone. This allows us to calibrate $\mathcal{J}$-invariant versions of any parameterised denoising algorithm, from the single hyperparameter of a median filter to the millions of weights of a deep neural network. We demonstrate this on natural image and microscopy data, where we exploit noise independence between pixels, and on single-cell gene expression data, where we exploit independence between detections of individual molecules. This framework generalizes recent work on training neural nets from noisy images and on cross-validation for matrix factorization.} }
Endnote
%0 Conference Paper %T Noise2Self: Blind Denoising by Self-Supervision %A Joshua Batson %A Loic Royer %B Proceedings of the 36th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Ruslan Salakhutdinov %F pmlr-v97-batson19a %I PMLR %P 524--533 %U https://proceedings.mlr.press/v97/batson19a.html %V 97 %X We propose a general framework for denoising high-dimensional measurements which requires no prior on the signal, no estimate of the noise, and no clean training data. The only assumption is that the noise exhibits statistical independence across different dimensions of the measurement, while the true signal exhibits some correlation. For a broad class of functions (“$\mathcal{J}$-invariant”), it is then possible to estimate the performance of a denoiser from noisy data alone. This allows us to calibrate $\mathcal{J}$-invariant versions of any parameterised denoising algorithm, from the single hyperparameter of a median filter to the millions of weights of a deep neural network. We demonstrate this on natural image and microscopy data, where we exploit noise independence between pixels, and on single-cell gene expression data, where we exploit independence between detections of individual molecules. This framework generalizes recent work on training neural nets from noisy images and on cross-validation for matrix factorization.
APA
Batson, J. & Royer, L.. (2019). Noise2Self: Blind Denoising by Self-Supervision. Proceedings of the 36th International Conference on Machine Learning, in Proceedings of Machine Learning Research 97:524-533 Available from https://proceedings.mlr.press/v97/batson19a.html.

Related Material