Contamination Map Methodology

Vision / Dataset

Real-IAD

Real-world, multi-view industrial anomaly detection dataset: about 150K high-resolution images (151,050 in the paper's comparison table) of 30 objects, each captured from 5 viewpoints, with pixel-level masks for anomalous images. Benchmark protocols: Unsupervised IAD (training set contains only normal samples; test set contains normal and anomalous samples) and Fully Unsupervised IAD (0-40% anomalous images added to the training set). Distributed as realiad_1024 (1024x1024), realiad_raw and JSON split files.

Also known as
Real-IAD dataset, RealIAD
Domain
manufacturing inspection
License
CC BY-NC-SA 4.0 (Hugging Face dataset card; released for research purpose only)
Test items from
2024-03-19
Evidence
“arXiv:2403.12580v1 [cs.CV] 19 Mar 2024”
arxiv.org · paper · header · retrieved 2026-10-15
Homepage
realiad4ad.github.io

Exposure by model

BenchmarkModelLevelReason
Industrial anomaly detection AA-CLIP, trained on MVTec AD (for VisA results) Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection AA-CLIP, trained on VisA Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection AdaCLIP, trained on all 14 datasets (demo model) Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection AdaCLIP, trained on MVTec AD + ColonDB Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection AdaCLIP, trained on VisA + ClinicDB Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection AdaptCLIP, trained on MVTec AD Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection AdaptCLIP, trained on VisA Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection Anomaly-OneVision (Anomaly-OV) Unknown unknown for its base model LLaVA-OneVision: training data not disclosed in enough detail
Industrial anomaly detection AnomalyCLIP, trained on MVTec AD Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection AnomalyCLIP, trained on VisA Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection AnomalyDINO Unknown unknown for its base model DINOv2: training data not disclosed in enough detail
Industrial anomaly detection AnomalyGPT, unsupervised on MVTec AD Unknown training data not disclosed in enough detail
Industrial anomaly detection AnomalyGPT, supervised (demo model) Unknown training data not disclosed in enough detail
Industrial anomaly detection AnomalyGPT, unsupervised on VisA Unknown training data not disclosed in enough detail
Industrial anomaly detection APRIL-GAN (VAND 2023), trained on MVTec AD Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection APRIL-GAN (VAND 2023), trained on VisA Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection OpenAI CLIP Unknown training data not disclosed in enough detail
Industrial anomaly detection DINOv2 Unknown training data not disclosed in enough detail
Industrial anomaly detection DINOv3 (web / LVD-1689M models) Unknown training data not disclosed in enough detail
Industrial anomaly detection FiLo, trained on MVTec AD Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection FiLo, trained on VisA Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection WideResNet-50-2 (supervised ImageNet-1k, torchvision) Not seen absent from the complete list of training data
Industrial anomaly detection InCTRL, trained on MVTec AD Not seen absent from the complete list of training data and from its base models' data
Industrial anomaly detection InCTRL, trained on VisA Not seen absent from the complete list of training data and from its base models' data
Industrial anomaly detection MetaUAS Unknown training data not disclosed in enough detail
Industrial anomaly detection MuSc Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection OpenCLIP (LAION-2B checkpoints) Not seen its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-01-31)
Industrial anomaly detection OpenCLIP (LAION-400M checkpoints) Not seen its test items were published from 2024-03-19, after the model's stated training-data cutoff (2021-12-31)
Industrial anomaly detection Segment Any Anomaly (SAA / SAA+) Unknown training data not disclosed in enough detail
Industrial anomaly detection SAM 2 / SAM 2.1 Unknown training data not disclosed in enough detail
Industrial anomaly detection Segment Anything Model (SAM) Not seen its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-12-31)
Industrial anomaly detection SigLIP 2 Not seen its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-12-31)
Industrial anomaly detection SigLIP Not seen its test items were published from 2024-03-19, after the model's stated training-data cutoff (2022-12-31)
Industrial anomaly detection UniVAD Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection VCP-CLIP, trained on MVTec-AD Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection VCP-CLIP, trained on VisA Unknown unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Industrial anomaly detection WinCLIP / WinCLIP+ Not seen absent from the complete list of training data and from its base models' data

Notes

published_from = arXiv v1 date (earliest documented); the dataset is distributed through a gated (auto-approved) Hugging Face repo whose API reports creation on 2024-04-11, so the files may have become downloadable a few weeks after the paper. Exact train/test image counts for the UIAD split come from the JSON split files and were not checked. Later companion datasets (Real-IAD D3, Real-IAD Variety) are separate and not covered here.

Sources