Estimating the probability density function of new fabrics for fabric anomaly detection
Rippel O, Müller M, Münkel A, Gries T, Merhof D (2021)
Publication Type: Conference contribution
Publication year: 2021
Publisher: SciTePress
Pages Range: 463-470
Conference Proceedings Title: ICPRAM 2021 - Proceedings of the 10th International Conference on Pattern Recognition Applications and Methods
Event location: Virtual, Online
ISBN: 9789897584862
Abstract
Image-based quality control aims at detecting anomalies (i.e. defects) in products. Supervised, data driven approaches have greatly improved Anomaly Detection (AD) performance, but suffer from a major drawback: they require large amounts of annotated training data, limiting their economic viability. In this work, we challenge and overcome this limitation for complex patterned fabrics. Investigating the structure of deep feature representations learned on a large-scale fabric dataset, we find that fabrics form clusters according to their fabric type, whereas anomalies form a cluster on their own. We leverage this clustering behavior to estimate the Probability Density Function (PDF) of new, previously unseen fabrics, in the deep feature representations directly. Using this approach, we outperform supervised and semi-supervised AD approaches trained on new fabrics, requiring only defect-free data for PDF-estimation.
Involved external institutions
Germany (DE)How to cite
APA:
Rippel, O., Müller, M., Münkel, A., Gries, T., & Merhof, D. (2021). Estimating the probability density function of new fabrics for fabric anomaly detection. In Maria De Marsico, Gabriella Sanniti di Baja, Ana Fred (Eds.), ICPRAM 2021 - Proceedings of the 10th International Conference on Pattern Recognition Applications and Methods (pp. 463-470). Virtual, Online: SciTePress.
MLA:
Rippel, Oliver, et al. "Estimating the probability density function of new fabrics for fabric anomaly detection." Proceedings of the 10th International Conference on Pattern Recognition Applications and Methods, ICPRAM 2021, Virtual, Online Ed. Maria De Marsico, Gabriella Sanniti di Baja, Ana Fred, SciTePress, 2021. 463-470.
BibTeX: Download