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

Anomaly Detection

1,150 평가 기록 172 datasets

Anomaly Detection is a binary classification identifying unusual or unexpected patterns in a dataset, which deviate significantly from the majority of the data. The goal of anomaly detection is to identify such anomalies, which could represent errors, fraud, or other types of unusual events, and flag them for further investigation.

[Image source]: GAN-based Anomaly Detection in Imbalance Problems

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
GLASS
Detection AUROC: 99.9Segmentation AUROC: 99.3Segmentation AUPRO: 96.8
논문 코드
UniNet
Detection AUROC: 99.8Segmentation AUPRO (until 30% FPR): 93.9Segmentation AUROC: 98.8Segmentation AUPRO: 93.9
논문 코드
CSAD
Avg. Detection AUROC: 95.3Detection AUROC (only logical): 96.7Detection AUROC (only structural): 94.0
논문 코드
CLIP (OE)
AUROC: 99.6
논문 코드
MULDE-object-centric-micro
AUC: 94.3%
논문 코드
DAC(STG-NF + Jigsaw)
AUC: 87.72%
논문 코드
Auto-Encoder with Regression (AER)
Average F1: 0.470
논문 코드
STEAD-Base
ROC AUC: 91.34
논문 코드
DUS-Net
Object AUROC: 0.795Point AUROC: 0.847Mean Performance of P. and O. : 0.821
논문 -
cDNP+OE
AP: 69.8FPR95: 7.5
논문 코드
CFA-HLGAtt
AP: 86.34
논문 -
GLASS
Detection AUROC: 99.6Segmentation AUROC: 99.4Segmentation AUPRO: 98.2
논문 코드

데이터셋

172개
MVTec ADVisAMVTec LOCO ADOne-class CIFAR-10CUHK AvenueShanghaiTechUCR Anomaly ArchiveUCF-CrimeReal 3D-ADFishyscapes L&FXD-ViolenceMPDD

논문

라이브러리

Papers with Code 리더보드 row에 연결된 구현 저장소입니다.

github.com/liznerski/eoe 1 논문
github.com/2023-MindSpore-1/ms-code-215/tree/main/PaDiM 1 논문
github.com/A03ki/f-AnoGAN 1 논문
github.com/advanced-vision-and-learning-lab/padim-ace 1 논문
github.com/Burf/tfdetection 1 논문
github.com/code-implementation1/Code6/tree/main/PaDiM 1 논문
github.com/CuberrChen/PaDiM-Paddle 1 논문
github.com/DannisZgggg/AnoGAN-WGAN-pytorch 1 논문

하위 태스크

3A 3D Anomaly Detection 21 데이터셋 · 130 rows IM Image Manipulation Detection 21 데이터셋 · 106 rows UA Unsupervised Anomaly Detection 23 데이터셋 · 78 rows AD Anomaly Detection In Surveillance Videos 7 데이터셋 · 62 rows AE Abnormal Event Detection In Video 5 데이터셋 · 23 rows AC Anomaly Classification 6 데이터셋 · 17 rows MA Multi-class Anomaly Detection 3 데이터셋 · 14 rows DA Depth Anomaly Detection and Segmentation 1 데이터셋 · 13 rows 3A 3D Anomaly Detection and Segmentation 2 데이터셋 · 12 rows SA Supervised Anomaly Detection 2 데이터셋 · 10 rows RA RGB+3D Anomaly Detection and Segmentation 1 데이터셋 · 9 rows CA Cold-Start Anomaly Detection 1 데이터셋 · 3 rows