Simon Thomine

AI for Pathology 🔬🧬🤖


2024 -
Senior computer vision research engineer at Vitadx. I develop AI for pathology to assist medical professionals in early cancer detection. Our primary focus is bladder cancer, the 7th most common cancer in men, along with research into blood and thyroid cancers. By leveraging advanced computer vision and multimodal AI, we aim to enhance diagnostic accuracy and improve patient outcomes.
2021 - 2024
PhD in deep learning from UTT and Aquilae, specializing in unsupervised learning for anomaly detection in industrial applications. Led multiple research and development projects at Aquilae, focusing on industry and security solutions.
2019 - 2020
Semester abroad at Laval University in Quebec, Canada, focusing on master-level courses in deep learning and applied mathematics.
2017 - 2020
Engineering degree in computer vision application development from Telecom Saint-Etienne.
2015 - 2017
Preparatory classes for engineering schools at ENCPB, Paris.
Projects
Deep Learning Course in French aims to teach deep learning from scratch, requiring only basic Python and math knowledge. It offers a comprehensive introduction to core deep learning principles within a single GitHub repository with various notebook lessons. Making AI accessible to a wider audience is crucial in today’s world, given AI's growing presence in our daily lives.
3rd Place Cytologia Data Challenge. I participated in the Cytologia Data Challenge to contribute to advancing hematological diagnostics through AI. My goal was to detect and classify white blood cells in images from blood smears, and I am proud to have secured a third-place finish. ,winning a prize of 5000€.
Industrial Defect Generator a new library dessigned to generate synthetic industrial defects based on proposed methods from the litterature.
Distillation Industrial Anomaly Detection regroups advanced methods utilizing knowledge distillation for industrial unsupervised anomaly detection.
Industrial textile dataset introduces a new fabric anomaly dataset with 10 textile categories captured using high-resolution factory cameras.
Implementation of the paper "Remembering Normality: Memory-guided Knowledge Distillation for Unsupervised Anomaly Detection".
Publications
Visual Anomaly Detection on Surfaces through Unsupervised Learning for the Future Industry
PhD Thesis 2024
Simon Thomine
Textile Research Journal 2024
Simon Thomine and Hichem Snoussi
VISAPP 2024
Simon Thomine and Hichem Snoussi
Pattern Analysis and Applications 2024
Simon Thomine and Hichem Snoussi
Textile Research Journal 2024
Simon Thomine and Hichem Snoussi
ICCAD 2023
Simon Thomine, Hichem Snoussi and Mahmoud Soua
VISAPP 2023
Simon Thomine, Hichem Snoussi and Mahmoud Soua