Papers with Code · CC BY-SA 4.0

Source-Free Domain Adaptation

Domain Adaptation › Source-Free Domain Adaptation 41 평가 기록 15 datasets

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

개요는 각 항목의 일부입니다. 위 탭에서 전체 기록을 페이지별로 탐색할 수 있습니다.

벤치마크

Papers with Code 아카이브의 데이터셋별 기록 예시입니다. 평가 조건과 지표 방향이 검증되지 않아 순위를 의미하지 않습니다.

Dataset기록된 모델 예시MetricPaperCode

Cityscapes to Foggy Cityscapes

GT

AP50: 50.8

논문 코드

VisDA-2017

RCL

Accuracy: 93.2

논문 코드

InBreast

GT

R@0.05: 0.06R@0.3: 0.45R@0.5: 0.65R@1.0: 0.92

논문 코드

PACS

SPM

Average Accuracy: 86.7

논문 코드

Cityscapes to ACDC

CMA

mIoU: 69.1

논문 코드

Cityscapes to Dark Zurich

CMA

mIoU: 53.6

논문 코드

GTA5 to Cityscapes

HALO

mIoU: 73.3

논문 코드

SYNTHIA-to-Cityscapes

STPL

mIoU: 51.8

논문 코드

SynLiDAR-to-SemanticKITTI

TTYD

mIoU: 32.4

논문 코드

SynLiDAR-to-SemanticPOSS

TTYD

mIoU: 39.1

논문 코드

VIPER-to-Cityscapes

STPL

mIoU: 52.5

논문 코드

nuScenes-to-Pandaset

TTYD

mIoU: 65.7

논문 코드

데이터셋

15개

Cityscapes to Foggy Cityscapes VisDA-2017 InBreast PACS Cityscapes to ACDC Cityscapes to Dark Zurich GTA5 to Cityscapes SYNTHIA-to-Cityscapes SynLiDAR-to-SemanticKITTI SynLiDAR-to-SemanticPOSS VIPER-to-Cityscapes nuScenes-to-Pandaset

논문

SF(DA)$^2$: Source-free Domain Adaptation Through the Lens of Data Augmentation

2403.10834 2024-03-16 이 페이지의 평가 2건

Accuracy: 89.6

Shuffle PatchMix Augmentation with Confidence-Margin Weighted Pseudo-Labels for Enhanced Source-Free Domain Adaptation

2505.24216 2025-05-30 이 페이지의 평가 2건

Average Accuracy: 86.7

C-SFDA: A Curriculum Learning Aided Self-Training Framework for Efficient Source Free Domain Adaptation

2303.17132 2023-03-30 이 페이지의 평가 1건

Accuracy: 87.8

Cleaning Noisy Labels by Negative Ensemble Learning for Source-Free Unsupervised Domain Adaptation

2022-02-15 이 페이지의 평가 1건

Average Accuracy: 72.4

Contrastive Model Adaptation for Cross-Condition Robustness in Semantic Segmentation

2303.05194 2023-03-09 이 페이지의 평가 1건

mIoU: 53.6

Divide and Contrast: Source-free Domain Adaptation via Adaptive Contrastive Learning

2211.06612 2022-11-12 이 페이지의 평가 1건

Accuracy: 87.3

Effective Dual-Region Augmentation for Reduced Reliance on Large Amounts of Labeled Data

2504.13077 2025-04-17 이 페이지의 평가 1건

Average Accuracy: 84.0

Empowering Source-Free Domain Adaptation with MLLM-driven Curriculum Learning

2405.18376 2024-05-28 이 페이지의 평가 1건

Accuracy: 93.2

Source Data-absent Unsupervised Domain Adaptation through Hypothesis Transfer and Labeling Transfer

2012.07297 2020-12-14 이 페이지의 평가 1건

Accuracy: 87.3

Spatio-Temporal Pixel-Level Contrastive Learning-based Source-Free Domain Adaptation for Video Semantic Segmentation

2303.14361 2023-03-25 이 페이지의 평가 1건

mIoU: 51.8

라이브러리

Papers with Code 리더보드 row에 연결된 구현 저장소입니다.

github.com/fenghz/cosda 2 논문

github.com/PrasannaPulakurthi/SPM 1 논문

github.com/shinyflight/sfda2 1 논문

github.com/brdav/cma 1 논문

github.com/Dong-Jie-Chen/RCL 1 논문

github.com/nazmul-karim170/C-SFDA_Source-Free-Domain-Adaptation 1 논문

github.com/PrasannaPulakurthi/Foreground-Background-Augmentation 1 논문

github.com/shaoyuanlo/stpl 1 논문

하위 태스크

SF Source Free Object Detection 2 데이터셋 · 16 rows 3S 3D Source-Free Domain Adaptation 6 데이터셋 · 6 rows