2022 - 2025
Currently pursuing a Master's degree in Data Science at EPFL, improving my understanding of research and scientific fundamentals in the field of AI and machine learning.
Award - Hopefully one 🤞
Grade - 5.56/6.00
2020 - 2022
Completed a Bachelor's degree in Computer Science with a specialization in Software Development and Imaging at HES-SO Arc Ingéniérie, which gave me a deep understanding of software development, web/app/game technologies, as well as image processing and computer vision.
Award - Top Student of the entire Engineering Field
Grade - 5.73/6.00
2019 - 2022
Completed a Bachelor's degree in Engineering at UTBM. This program equipped me with foundational skills in Engineering and Computer Science.
Award - DEUTEC Prize for the Best Students
Grade - 5.06/6.00
Autumn 2024
Developed a machine learning pipeline to generate pre-annotations for spinal root segmentation from MRI scans, supporting neuroprosthetic solutions for Spinal Cord Injury (SCI) patients. Leveraged nnU-Net for adaptable segmentation and HarDNet for efficient processing, addressing challenges such as limited data and closely packed spinal structures. Enhanced pre-annotation quality through advanced pre-processing, augmentation, and post-processing, demonstrating how AI can streamline personalized spinal cord reconstruction and reduce manual effort in clinical workflows.
Based on nnUNet and HarDNet
Spring - Autumn 2024
Pre-trained a 3D Masked Autoencoder based on Vision Transformers (ViT) to learn the underlying structure and language of the spinal cord from MRI data. This foundational model was fine-tuned for downstream tasks such as spinal cord segmentation and artefact removal, achieving improved performance and efficiency. The pre-training significantly enhanced fine-tuning on these tasks, reducing the need for large, annotated datasets, which are often costly and time-consuming to obtain in the medical field.
Based on META Masked Autoencoders
Spring 2024
Developed an AI tutor using the GPT-2 model to assist students with scientific courses by providing explanations, answering questions, and offering personalized support. We fine-tuned the model using techniques like Simple Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Reinforcement Learning from Human Feedback (RLHF). This project showcased the potential of NLP to make complex scientific topics more accessible and engaging for students.
Based on OpenAI GPT2
Autumn 2023
Developed a deep learning model to decode arm movement intentions from ECoG brain recordings, aimed at aiding movement recovery for Spinal Cord Injury (SCI) patients. Using CNN-LSTM, EEGNet, and CEBRA-based models, we significantly improved movement classification. This demonstrated how deep learning enhances motor function interpretation, showcasing its potential in movement recovery.
Based on EEGNet and CEBRA
Summer 2022
Developed an advanced autopilot and target tracking system for JetRacer cars, based on NVIDIA's Jetson-Inference project and the Single-Shot Multiple Detection (SSD) algorithm. Using object detection and computer vision techniques, we trained the system to recognize and follow a target placed on the back of a controlled car, significantly enhancing its autonomous capabilities. This project demonstrated the potential of AI to improve real-time object tracking and autonomous navigation.
Based on NVIDIA Jetson-Inference
2022 - 2025
Oversaw IT operations and improved the website and application, focusing on enhancing the user experience. I also participated in organizing the event and ensuring the well-being of participants. On the event day, I led the coordination of the volleyball tournament, the biggest sport of the event, ensuring everything ran smoothly.
2020 - 2022
Guided a group of youths through weekly meetings, helping them develop communication, leadership, and conflict resolution skills. We focused on building their autonomy through activities like team projects and community service. During summer camp, we went camping and lived in nature, giving them a chance to apply these skills in a real-world setting.
