Informatics at the service of collective well-being and personal development.
The Faculty of Informatics has a vision of a world in which the intensive and growing use of informatics is oriented towards collective well-being and personal development. Its mission is to contribute to this orientation through education, scientific research and service to society.
The studies
When you choose to study computer science, you're guaranteed immediate access to an exciting, multi-faceted career. Through a comprehensive range of courses - bachelor's, master's, specialization master's and doctorate - the Faculty of Computer Science offers you quality training based on scientific rigor and critical thinking, while emphasizing interdisciplinarity and societal responsibility.
Research
The mission of the Faculty of Computer Science is to ensure cutting-edge scientific research, open to the creation and integration of knowledge that feeds teaching, technological innovation and societal debate.
Service to society
The Faculty of Computer Science contributes to the development of our social, economic, technological and political environment by disseminating knowledge and providing advice at regional, national and international levels. Through the synergy between researchers and industry, and by making knowledge and know-how available, it participates in numerous missions of service to society.
International
The Faculty of Computer Science's international outlook is fundamental to the development of its teaching and research activities.
Organization
The Faculty of Computer Science has organized itself to manage its missions in the best possible way. It has around a hundred members at the service of teaching, research and service to society.
A word from the Dean
IT is a key to understanding, progress and responsibility in the face of the challenges and issues facing our society.
Animation
A series of events and players revolve around the Faculty of Computer Science.
Spotlight
Agenda
Public defense of doctoral thesis - Arnaud BOUGAHAM
Generative Models and Quality Constraints for Anomaly Detection: Application to Industrial and Medical Images
Jury
- Prof. Benoit Frénay, University of Namur (Promoter)
- Prof. Isabelle Linden, University of Namur (Co-Promoter)
- Prof. Xavier Devroey, University of Namur (President)
- Prof. Claire Lobet, University of Namur
- Prof. Florentin Kucharczak, University of Montpellier
- Mr Olivier Dequenne
- Mr Achimi Hassaini
Abstract
This thesis proposes a robust, interpretable, and transferable deep-learning framework for anomaly detection in safety-critical domains such as industrial quality control and medical diagnostics. These two fields, though distinct, share major challenges: class imbalance with limited abnormal samples, and the need for trustworthy and real-time decisions under strict reliability constraints. The main objective is, thus, to build methods with minimal supervision (training with normal data only) while allowing a human-aligned interpretability.
The approach combines artificial intelligence unsupervised generative modeling with supervised classification, focusing on patching techniques, local representation, and interpretable scoring. Four key contributions structure this work:
(i) Generative Adversarial Network Anomaly Detection through Intermediate
Patches (GanoDIP), a Generative Adversarial Network (GAN) architecture for high-resolution, industrial anomaly localization at the patch level.
(ii) Vector Quantized Generative Adversarial Network Anomaly Detection
through Intermediate Patches (VQGanoDIP), an extension with vector-quantized latent representations and composite scoring for improved reconstruction and fidelity.
(iii) Cycle Generative Adversarial Network-Anomaly Detection (CGAN-AD), a conditional image translation model that integrates both normal and abnormal data for enhanced domain transfer in industrial and medical settings.
(iv) Trustworthy approximated partial AUC (tapAUC), a loss function that enforces the Zero False Negative constraint, for high recall in critical scenarios.
These models are deployed in real-world use cases. The framework is integrated into an active production line to detect unexpected components (such as screws) in printed circuit boards, delivering interpretable decisions with minimal false alarms. In the medical domain, the approach is adapted to well perform in Positron Emission Tomography (PET) based coma receptivity analysis, and some techniques are incorporated for an ovarian cancer segmentation application. These results demonstrate not only technical efficiency but also organizational viability through human-in-the-loop deployment and real-time scalability.
Together, these contributions establish a modular, constraint-aware, and explainable anomaly detection paradigm, advancing the field towards trustworthy and human-centered artificial intelligence in high-stakes environments.