CoinDefect6k: A Large-Scale Industrial Dataset for Multi-Class Visual Anomaly Detection

This Zenodo record provides a curated image dataset of coins acquired on a production-grade high-speed inspection line (Proditec VISIA) and annotated for industrial visual anomaly detection and defect-type classification. Images originate from two sources ("dataset families"): Manufacturing defects: defects introduced by manufacturing/sorting settings and typical production issues (e.g., partially erased design elements, pressure inconsistencies). Circulation quality: defects associated with realistic wear and damage during real-world circulation (e.g., oxidation, dirt, physical damage). The datset was collected in collaboratioon with Proditec company wihtin PEER project. This work is part of a project that has received funding from the European Union's Horizon Europe Research and Innovation Programme, under Grant Agreement number 101120406. The paper reflects only the authors' view and the EC is not responsible for any use that may be made of the information it contains.