SOTAVerified

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 AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 76–100 of 188 papers

TitleStatusHype
Local-Global Pseudo-label Correction for Source-free Domain Adaptive Medical Image Segmentation—0
Domain-Specificity Inducing Transformers for Source-Free Domain Adaptation—0
Prior-guided Source-free Domain Adaptation for Human Pose Estimation—0
Unsupervised Adaptation of Polyp Segmentation Models via Coarse-to-Fine Self-Supervision—0
Cal-SFDA: Source-Free Domain-adaptive Semantic Segmentation with Differentiable Expected Calibration ErrorCode1
ReCLIP: Refine Contrastive Language Image Pre-Training with Source Free Domain AdaptationCode1
Consistency Regularization for Generalizable Source-free Domain Adaptation—0
Divide and Adapt: Active Domain Adaptation via Customized LearningCode1
Feed-Forward Source-Free Domain Adaptation via Class Prototypes—0
Source-Free Domain Adaptation for Medical Image Segmentation via Prototype-Anchored Feature Alignment and Contrastive LearningCode1
Unsupervised Accuracy Estimation of Deep Visual Models using Domain-Adaptive Adversarial Perturbation without Source Samples—0
Source-Free Domain Adaptive Fundus Image Segmentation with Class-Balanced Mean TeacherCode1
Source-Free Domain Adaptation with Temporal Imputation for Time Series DataCode1
Generating Reliable Pixel-Level Labels for Source Free Domain Adaptation—0
Learning Content-enhanced Mask Transformer for Domain Generalized Urban-Scene SegmentationCode0
Hyperbolic Active Learning for Semantic Segmentation under Domain ShiftCode1
Towards Source-free Domain Adaptive Semantic Segmentation via Importance-aware and Prototype-contrast LearningCode1
FACT: Federated Adversarial Cross TrainingCode0
Source-Free Domain Adaptation for SSVEP-based Brain-Computer InterfacesCode0
Source-Free Domain Adaptation for RGB-D Semantic Segmentation with Vision Transformers—0
Imbalance-Agnostic Source-Free Domain Adaptation via Avatar Prototype Alignment—0
Black-box Source-free Domain Adaptation via Two-stage Knowledge Distillation—0
CoSDA: Continual Source-Free Domain AdaptationCode1
Source-free Domain Adaptation Requires Penalized Diversity—0
Few-shot Fine-tuning is All You Need for Source-free Domain AdaptationCode0
Show:102550
← PrevPage 4 of 8Next →
All datasetsVisDA-2017PACSCityscapes to ACDCCityscapes to Dark Zurich

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1RCLAccuracy93.2—Unverified
2SFDA2++Accuracy89.6—Unverified
3SPMAccuracy89.4—Unverified
4SFDA2Accuracy88.1—Unverified
5C-SFDAAccuracy87.8—Unverified
6DaCAccuracy87.3—Unverified
7SHOT++Accuracy87.3—Unverified
8NRCAccuracy85.9—Unverified
9G-SFDAAccuracy85.4—Unverified
10SHOTAccuracy82.9—Unverified
#ModelMetricClaimedVerifiedStatus
1SPMAverage Accuracy86.7—Unverified
2DRAAverage Accuracy84—Unverified
3NELAverage Accuracy72.4—Unverified
#ModelMetricClaimedVerifiedStatus
1CMAmIoU69.1—Unverified
#ModelMetricClaimedVerifiedStatus
1CMAmIoU53.6—Unverified