Probabilistic Noise2Void: Unsupervised Content-Aware Denoising
| dc.contributor.author | Krull, Alexander | cs |
| dc.contributor.author | Vičar, Tomáš | cs |
| dc.contributor.author | Prakash, Mangal | cs |
| dc.contributor.author | Lalit, Manan | cs |
| dc.contributor.author | Jug, Florian | cs |
| dc.coverage.issue | 5 | cs |
| dc.coverage.volume | 2 | cs |
| dc.date.accessioned | 2020-07-29T06:57:01Z | |
| dc.date.available | 2020-07-29T06:57:01Z | |
| dc.date.issued | 2020-02-19 | cs |
| dc.description.abstract | Today, Convolutional Neural Networks (CNNs) are the leading method for image denoising. They are traditionally trained on pairs of images, which are often hard to obtain for practical applications. This motivates self-supervised training methods, such as Noise2Void (N2V) that operate on single noisy images. Self-supervised methods are, unfortunately, not competitive with models trained on image pairs. Here, we present Probabilistic Noise2Void (PN2V), a method to train CNNs to predict per-pixel intensity distributions. Combining these with a suitable description of the noise, we obtain a complete probabilistic model for the noisy observations and true signal in every pixel. We evaluate PN2V on publicly available microscopy datasets, under a broad range of noise regimes, and achieve competitive results with respect to supervised state-of-the-art methods. | en |
| dc.format | text | cs |
| dc.format.extent | 1-9 | cs |
| dc.format.mimetype | application/pdf | cs |
| dc.identifier.citation | Frontiers in Computer Science. 2020, vol. 2, issue 5, p. 1-9. | en |
| dc.identifier.doi | 10.3389/fcomp.2020.00005 | cs |
| dc.identifier.issn | 2624-9898 | cs |
| dc.identifier.other | 159778 | cs |
| dc.identifier.uri | http://hdl.handle.net/11012/193231 | |
| dc.language.iso | en | cs |
| dc.publisher | Frontiers Media SA | cs |
| dc.relation.ispartof | Frontiers in Computer Science | cs |
| dc.relation.uri | https://www.frontiersin.org/articles/10.3389/fcomp.2020.00005/full | cs |
| dc.rights | Creative Commons Attribution 4.0 International | cs |
| dc.rights.access | openAccess | cs |
| dc.rights.sherpa | http://www.sherpa.ac.uk/romeo/issn/2624-9898/ | cs |
| dc.rights.uri | http://creativecommons.org/licenses/by/4.0/ | cs |
| dc.subject | denoising | en |
| dc.subject | CARE | en |
| dc.subject | deep learning | en |
| dc.subject | microscopy data | en |
| dc.subject | probabilistic | en |
| dc.title | Probabilistic Noise2Void: Unsupervised Content-Aware Denoising | en |
| dc.type.driver | article | en |
| dc.type.status | Peer-reviewed | en |
| dc.type.version | publishedVersion | en |
| sync.item.dbid | VAV-159778 | en |
| sync.item.dbtype | VAV | en |
| sync.item.insts | 2023.03.13 12:52:49 | en |
| sync.item.modts | 2023.03.13 12:14:34 | en |
| thesis.grantor | Vysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií. Ústav biomedicínského inženýrství | cs |
Files
Original bundle
1 - 1 of 1
Loading...
- Name:
- fcomp0200005.pdf
- Size:
- 2.96 MB
- Format:
- Adobe Portable Document Format
- Description:
- fcomp0200005.pdf