Computer Vision Researcher @ Idemia · Télécom Paris & ENS Paris-Saclay
Master in Mathematics, Vision, Learning (MVA)
Leading Mathematics, Vision, Learning algorithm Master.
Engineering Degree - GPA: 4/4
Specializations: Image Processing, Signal Processing, Artificial Intelligence.
Preparatory Classes for Grandes Écoles (CPGE), MPSI-MP*
Idemia Public Security — Courbevoie, France
Literature review on object, person and luggage re-identification (Re-ID). Design and implementation of a baseline architecture for luggage Re-ID, followed by multi-view and 3D modules achieving state-of-the-art performance.
ANFR (Agence Nationale des Fréquences) — Paris
Frequency spectrum monitoring and equipment labelling for the Paris 2024 Olympic and Paralympic Games.
Freelance
Private tutoring in mathematics and physics for middle school, high school and preparatory cycle students.
Implementation of NCSN and Diffusion models. Evaluation of generation and inpainting capabilities on the MedMNIST benchmark.
Development of a multimodal Deep Learning framework (GCN + SciBERT) with contrastive loss to align molecular graph embeddings and textual descriptions.
Implementation of SimCLR and Barlow Twins methods. Evaluation of self-supervised learning effectiveness on the MedMNIST dataset.
Fraudulent transaction classification on a 300k-row DataFrame (Random Forest, feature engineering). Ranked 3rd out of 25 teams (F1 score: 69%).
Development of a sign language recognition prototype using the MediaPipe library and training of a Random Forest model.
Algorithmic implementation of the method detailed in the paper "Object Removal by Exemplar-Based Inpainting" (A. Criminisi et al.).
Assembly (Raspberry Pi, motors), GPIO configuration, and Python code design for remote control and automatic navigation via beacon detection.