Wienerberger
Roof Tile Anomaly Detection
Unsupervised computer-vision quality control that flags broken roof tiles on a high-volume factory line, trained from a sparse labelled dataset.
What we built
- Unsupervised anomaly detection model in Python and PyTorch
- Vision Transformer feature extraction with One-Class SVM scoring
- Defect detection tuned for limited labelled data on a production line
Delivered by Pavlo Holubchykov, Raw Code's Lead ML/AI Developer.
Outcome metrics for this project haven't been independently verified yet — figures will be added here once confirmed rather than estimated.
More work
Want something like this built?
Book a call