Keywords: source-free domain adaptation, segmentation
TL;DR: We tackle source-free domain adaptation (SFDA) for semantic segmentation.
Abstract: Domain adaptation (DA) tackles the performance drop observed when applying a model on target data from a different domain than the training one. However, most common DA techniques require concurrent access to the input images of both the source and target domains, which is often impossible for privacy concerns. We introduce a source-free domain adaptation for image segmentation, leveraging a prior-aware entropy minimization. We validate on spine, prostate and cardiac segmentation problems.
Our method yields comparable results to several state-of-the-art adaptation techniques, despite having access to much less information. Our framework can be used in many segmentation problems, and our code is publicly available at \url{https://github.com/mathilde-b/SFDA}
Registration: I acknowledge that acceptance of this work at MIDL requires at least one of the authors to register and present the work during the conference.
Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.
Paper Type: recently published or submitted journal contributions
Primary Subject Area: Transfer Learning and Domain Adaptation
Secondary Subject Area: Segmentation
Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.
1 Reply
Loading