Source-Free Domain Adaptation

Source-Free Domain Adaptation (SFDA) is a domain adaptation method in machine learning and computer vision where the goal is to adapt a pre-trained model to a new, target domain without access to the source domain data. This approach is advantageous in scenarios where sharing the source data is impractical due to privacy concerns, data size, or proprietary restrictions

Papers

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 11–20 of 188 papers

Title	Date	Tasks	Status	Hype
Simplifying Source-Free Domain Adaptation for Object Detection: Effective Self-Training Strategies and Performance Insights	Jul 10, 2024	Domain Adaptationobject-detection	CodeCode Available	1
An Uncertainty-guided Tiered Self-training Framework for Active Source-free Domain Adaptation in Prostate Segmentation	Jul 3, 2024	Domain AdaptationImage Segmentation	CodeCode Available	1
GLC++: Source-Free Universal Domain Adaptation through Global-Local Clustering and Contrastive Affinity Learning	Mar 21, 2024	ClusteringContrastive Learning	CodeCode Available	1
SF(DA)^2: Source-free Domain Adaptation Through the Lens of Data Augmentation	Mar 16, 2024	Data AugmentationDisentanglement	CodeCode Available	1
DG-TTA: Out-of-domain Medical Image Segmentation through Augmentation and Descriptor-driven Domain Generalization and Test-Time Adaptation	Dec 11, 2023	Domain AdaptationDomain Generalization	CodeCode Available	1
Robust Source-Free Domain Adaptation for Fundus Image Segmentation	Oct 25, 2023	Domain AdaptationImage Segmentation	CodeCode Available	1
UPL-SFDA: Uncertainty-aware Pseudo Label Guided Source-Free Domain Adaptation for Medical Image Segmentation	Sep 19, 2023	Brain SegmentationDomain Adaptation	CodeCode Available	1
Cal-SFDA: Source-Free Domain-adaptive Semantic Segmentation with Differentiable Expected Calibration Error	Aug 6, 2023	Domain AdaptationModel Selection	CodeCode Available	1
ReCLIP: Refine Contrastive Language Image Pre-Training with Source Free Domain Adaptation	Aug 4, 2023	Domain Adaptationimage-classification	CodeCode Available	1
Divide and Adapt: Active Domain Adaptation via Customized Learning	Jul 21, 2023	Active LearningDomain Adaptation	CodeCode Available	1

Show:10 25 50

← PrevPage 2 of 19Next →

All datasets VisDA-2017 PACS Cityscapes to ACDC Cityscapes to Dark Zurich

Benchmark Results

#	Model	Metric	Claimed	Verified	Status
1	RCL	Accuracy	93.2	—	Unverified
2	SFDA2++	Accuracy	89.6	—	Unverified
3	SPM	Accuracy	89.4	—	Unverified
4	SFDA2	Accuracy	88.1	—	Unverified
5	C-SFDA	Accuracy	87.8	—	Unverified
6	DaC	Accuracy	87.3	—	Unverified
7	SHOT++	Accuracy	87.3	—	Unverified
8	NRC	Accuracy	85.9	—	Unverified
9	G-SFDA	Accuracy	85.4	—	Unverified
10	SHOT	Accuracy	82.9	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	SPM	Average Accuracy	86.7	—	Unverified
2	DRA	Average Accuracy	84	—	Unverified
3	NEL	Average Accuracy	72.4	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	CMA	mIoU	69.1	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	CMA	mIoU	53.6	—	Unverified