Please use this identifier to cite or link to this item: https://scidar.kg.ac.rs/handle/123456789/15894
Title: Towards Measuring Domain Shift in Histopathological Stain Translation in an Unsupervised Manner
Authors: Nisar Z.
Vasiljević, Jelica
Ganarski P.
Lampert, Thomas
Issue Date: 2022
Abstract: Domain shift in digital histopathology can occur when different stains or scanners are used, during stain translation, etc. A deep neural network trained on source data may not generalise well to data that has undergone some domain shift. An important step towards being robust to domain shift is the ability to detect and measure it. This article demonstrates that the PixelCNN and domain shift metric can be used to detect and quantify domain shift in digital histopathology, and they demonstrate a strong correlation with generalisation performance. These findings pave the way for a mechanism to infer the average performance of a model (trained on source data) on unseen and unlabelled target data.
URI: https://scidar.kg.ac.rs/handle/123456789/15894
Type: conferenceObject
DOI: 10.1109/ISBI52829.2022.9761411
ISSN: 1945-7928
SCOPUS: 2-s2.0-85129605142
Appears in Collections:Faculty of Science, Kragujevac

Page views(s)

341

Downloads(s)

7

Files in This Item:
File Description SizeFormat 
PaperMissing.pdf
  Restricted Access
29.86 kBAdobe PDFThumbnail
View/Open


Items in SCIDAR are protected by copyright, with all rights reserved, unless otherwise indicated.