Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

DomainAdaptor: A Novel Approach to Test-time Adaptation

About

To deal with the domain shift between training and test samples, current methods have primarily focused on learning generalizable features during training and ignore the specificity of unseen samples that are also critical during the test. In this paper, we investigate a more challenging task that aims to adapt a trained CNN model to unseen domains during the test. To maximumly mine the information in the test data, we propose a unified method called DomainAdaptor for the test-time adaptation, which consists of an AdaMixBN module and a Generalized Entropy Minimization (GEM) loss. Specifically, AdaMixBN addresses the domain shift by adaptively fusing training and test statistics in the normalization layer via a dynamic mixture coefficient and a statistic transformation operation. To further enhance the adaptation ability of AdaMixBN, we design a GEM loss that extends the Entropy Minimization loss to better exploit the information in the test data. Extensive experiments show that DomainAdaptor consistently outperforms the state-of-the-art methods on four benchmarks. Furthermore, our method brings more remarkable improvement against existing methods on the few-data unseen domain. The code is available at https://github.com/koncle/DomainAdaptor.

Jian Zhang, Lei Qi, Yinghuan Shi, Yang Gao• 2023

Related benchmarks

TaskDatasetResultRank
Image ClassificationPACS
Overall Average Accuracy88.45
230
Multi-class classificationVLCS
Acc (Caltech)98.69
139
Domain GeneralizationDomainBed
Average Accuracy62.5
127
Prostate SegmentationHK (test)
DSC85.2
20
Polyp SegmentationPolyp Segmentation Site D (trained on Site C) (test)
DSC81.33
10
Optic Disc and Optic Cup SegmentationOD/OC (Domain D)
DSC56.78
10
Polyp SegmentationPolyp Site A (test)
DSC88.58
10
Polyp SegmentationPolyp Segmentation Site B (trained on Site A) (test)
DSC78.39
10
Polyp SegmentationPolyp Segmentation Site B (trained on Site C) (test)
DSC73.4
10
Polyp SegmentationPolyp Segmentation Site A (trained on Site D) (test)
DSC86.1
10
Showing 10 of 38 rows

Other info

Follow for update