The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.

BTAD (beanTech Anomaly Detection) — pointer card, not a data mirror

Metadata/citation card only. The source zip (avires.dimi.uniud.it/papers/btad/btad.zip) is too large to inspect for license terms without downloading, and no explicit license text was found associated with it — not re-hosted until confirmed.

Access

Contents

2,830 RGB images, 3 products, pixel-level anomaly masks (body/surface defects).

Citation

Mishra, P., Verk, R., Fornasier, D., Piciarelli, C., Foresti, G.L. "VT-ADL: A Vision Transformer Network for Image Anomaly Detection and Localization." ISIE 2021.

Action needed: download and inspect the zip directly (or contact the authors) to confirm license before mirroring.

Downloads last month
-