PK PyTorchKR Papers
Papers with Code · CC BY-SA 4.0

Domain Adaptation

1,518 평가 기록 170 datasets

Domain Adaptation is the task of adapting models across domains. This is motivated by the challenge where the test and training datasets fall from different data distributions due to some factor. Domain adaptation aims to build machine learning models that can be generalized into a target domain and dealing with the discrepancy across domain distributions.

Further readings:

<span style="color:grey; opacity: 0.6">( Image credit: Unsupervised Image-to-Image Translation Networks )</span>

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
SIMPLE+
Average Accuracy: 99.0
논문 코드
VOLO-D5
Accuracy - All Images: 57.2Accuracy - Corrupted Images: 51.8Accuracy - Clean Images: 59.7
논문 코드
DINOv2 (ViT-g/14, frozen model, linear eval)
mean Corruption Error (mCE): 28.2Number of params: 1100M
논문 코드
MoA (OpenCLIP, ViT-B/16)
Average Accuracy: 90.6
논문 코드
FFTAT
Average Accuracy: 96.0
논문 코드
Model soups (BASIC-L)
Top-1 accuracy %: 94.17
논문 코드
Model soups (BASIC-L)
Top-1 Error Rate: 3.90
논문 코드
L2C (CLIP, ViT-L/14)
Average Accuracy: 67.4
논문 코드
CAR-FT (CLIP, ViT-B/16)
Average Accuracy: 85.5
논문 -
HALO
mIoU: 78.1Extra Manual Annotation: Yes
논문 코드
UniDG + CORAL + ConvNeXt-B
Average Accuracy: 69.6
논문 코드
SWG
Accuracy: 92.3
논문 코드

데이터셋

170개
PACSVizWiz-ClassificationImageNet-COffice-HomeOffice-31ImageNet-AImageNet-RDomainNetVLCSSYNTHIA-to-CityscapesTerraIncognitaOffice-Home

논문

라이브러리

Papers with Code 리더보드 row에 연결된 구현 저장소입니다.

github.com/chen742/PiPa 2 논문
github.com/sohyun-l/fifo 2 논문
github.com/BiQiWHU/BWG 1 논문
github.com/bupt-ai-cz/Meta-SelfLearning 1 논문
github.com/chen742/DCF 1 논문
github.com/Cuzyoung/CoDA 1 논문
github.com/lhoyer/mic 1 논문
github.com/licongguan/ilm-assl 1 논문

하위 태스크

DG Domain Generalization 25 데이터셋 · 596 rows UD Unsupervised Domain Adaptation 49 데이터셋 · 393 rows UD Universal Domain Adaptation 4 데이터셋 · 51 rows SD Source-Free Domain Adaptation 15 데이터셋 · 41 rows PD Partial Domain Adaptation 5 데이터셋 · 28 rows BD Blended-target Domain Adaptation 3 데이터셋 · 6 rows CI Continuously Indexed Domain Adaptation 3 데이터셋 · 4 rows WU Wildly Unsupervised Domain Adaptation 4 데이터셋 · 4 rows OM Open-Set Multi-Target Domain Adaptation 3 데이터셋 · 3 rows TA Test-time Adaptation 1 데이터셋 · 1 rows OD Online Domain Adaptation 0 데이터셋 · 0 rows PZ Prompt-driven Zero-shot Domain Adaptation 0 데이터셋 · 0 rows