Fully Unsupervised Probabilistic Noise2Void


Mangal Prakash*
Manan Lalit*
Pavel Tomancak

Alex Krull+
Florian Jug+

Center for Systems Biology Dresden
Max-Planck Institute of Molecular Cell Biology and Genetics

* Equal Contribution
+ Joint Supervision

In ISBI 2020

[pdf]
[code]
[slides]


Our proposed GMM bootstrapping approach does not require paired training or calibration data, but achieves superior results compared to other fully unsupervised methods.

Image denoising is the first step in many biomedical image analysis pipelines and Deep Learning (DL) based methods are currently best performing. A new category of DL methods such as Noise2Void or Noise2Self can be used fully unsupervised, requiring nothing but the noisy data. However, this comes at the price of reduced reconstruction quality. The recently proposed Probabilistic Noise2Void (PN2V) improves results, but requires an additional noise model for which calibration data needs to be acquired. Here, we present improvements to PN2V that (i) replace histogram based noise models by parametric noise models, and (ii) show how suitable noise models can be created even in the absence of calibration data. This is a major step since it actually renders PN2V fully unsupervised. We demonstrate that all proposed improvements are not only academic but indeed relevant.

Overview

A visual comparison of results obtained by CARE, N2V, PN2V, and our proposed methods (bold). We distinguish three families of methods: fully supervised (CARE), unsupervised but requiring additional calibration data (PN2V, our PN2V GMM),and fully unsupervised (N2V, PN2V using our bootstrapped histogram and GMM based noise models). The leftmost column in the unsupervised + calibration category shows the average of all available calibration images used for PN2V and PN2V GMM. Note that results of our fully unsupervised methods reach very similar quality to methods requiring either clean ground truth, or additional calibration data.

Results

Comparision of the denoising performance of all tested methods. Mean PSNR and ±1 standard error over five repetitions of each experiment are shown. Names of our proposed methods are shown in bold. Bold numbers indicate the best performing method in its respective category (supervised,unsupervised + calibration, and fully unsupervised; from top to bottom, separated by dashed lines)

Paper

Prakash*, Lalit*, Tomancak, Krull+, Jug+.
Fully Unsupervised Probabilistic Noise2Void.
In ISBI 2020.

[pdf]
[bibtex]

Acknowledgements

The authors would like to acknowledge the Light Microscopy Facility of MPI-CBG, Diana Afonso and Jacqueline Tabler from MPI-CBG for kindly sharing their samples and expertise, and Matthias Arzt from CSBD/MPI-CBG for helpful discussions on possible noise model formulations. This work was supported by the German Federal Ministry of Research and Education (BMBF) under the codes 031L0102 (de.NBI)and 01IS18026C (ScaDS2), and the German Research Foundation (DFG) under the code JU3110/1-1(FiSS). This webpage template was borrowed from some colorful folks.