no code implementations • 25 Feb 2022 • Antoine Cordier, Pierre Gutierrez, Victoire Plessis
As a consequence, models trained under such constraints are expected to be very sensitive to input distribution changes, which may be caused in practice by changes in the acquisition system (cameras, lights), in the parts or in the defects aspect.
no code implementations • 25 Feb 2022 • Antoine Cordier, Benjamin Missaoui, Pierre Gutierrez
In this work, we first assess the robustness of these pre-trained methods to fully unsupervised context, using polluted training sets (i. e. containing defective samples), and show that these methods are more robust to pollution compared to methods such as CutPaste.
no code implementations • 2 Jun 2021 • Pierre Gutierrez, Antoine Cordier, Thaïs Caldeira, Théophile Sautory
The use of deep features coming from pre-trained neural networks for unsupervised anomaly detection purposes has recently gathered momentum in the computer vision field.
no code implementations • 7 Apr 2021 • Pierre Gutierrez, Maria Luschkova, Antoine Cordier, Mustafa Shukor, Mona Schappert, Tim Dahmen
In order to detect defects, supervised learning is often utilized, but necessitates a large amount of annotated images, which can be costly: collecting, cleaning, and annotating the data is tedious and limits the speed at which a system can be deployed as everything the system must detect needs to be observed first.
no code implementations • 7 Apr 2021 • Antoine Cordier, Deepan Das, Pierre Gutierrez
In this work, we develop a methodology for learning actively, from rapidly mined, weakly (i. e. partially) annotated data, enabling a fast, direct feedback from the operators on the production line and tackling a big machine vision weakness: false positives.