Document is current
Any future updates will be listed below
-
Authors
Rolih, Blaž https://orcid.org/0009-0007-8630-3502
Fučka, Matic https://orcid.org/0009-0007-9654-8386
Skočaj, Danijel https://orcid.org/0000-0002-5290-4736
-
Funding
Funding for this research was provided by:
Javna Agencija za Raziskovalno Dejavnost RS (L2-3169, J2-60055, P2-0214, GC-0001)
- License Information
-
More Information
Article History
Received: 1 March 2025
Accepted: 21 August 2025
First Online: 23 September 2025
Declarations
:
: The authors have no Conflict of interest to declare that are relevant to the content of this article.
: This article is a substantially expanded version of our conference paper, titled SuperSimpleNet: Unifying Unsupervised and Supervised Learning for Fast and Reliable Surface Defect Detection published at International Conference on Pattern Recognition. In the conference version, we extended SimpleNet to also operate in fully supervised mode. In the manuscript that we are submitting here, we updated the architecture and further extended the method for two more supervision regimes: weakly and mixed. This is accompanied by extensive new evaluation and discussion. To comply with the double-blind review process, we didn’t cite the conference paper, but included it as a footnote in this final copy.
