Dec 2022 – Jul 2024 · 1 yr 7 mos · Stanford, California, United States

• Research in machine learning for 3D pose estimation and human motion as part of the OpenCap project of Stanford University.

Mar 2022 – Aug 2022 · 5 mos · Santa Monica, Californie, États-Unis

• Polygon is a new kind of psychology practice specializing in online testing and support for dyslexia, ADHD, and other learning differences. We utilize technology to provide a better experience for both clients and our team of in-house licensed psychologists.
• Initiated the data aggregation pipeline of the company.
• Produced a literature review on face detection, facial landmarks and head pose prediction using machine learning techniques (CNN, Visual Transformers, U-Net, Stacked Hourglass Networks).
• Implemented state-of-the-art models in face detection and landmarks prediction with Pytorch.
• Conducted exploratory work on 2D-to-3D point clouds prediction for facial landmarks.
• Implemented an active learning pipeline to improve facial landmarks prediction models with Pytorch.

Aug 2021 – Mar 2022 · 7 mos

• Digital Tech Year program of CentraleSupélec. This program consists of 6 months prototyping 3 Minimum Viable Products using agile methodology for different companies seeking help in Data Science and software engineering projects.

May 2021 – Aug 2021 · 3 mos

• Participation to the IEEE ICAS Challenge for anomaly detection in heterogeneous multi-robot systems.
• Led the university team to the 3rd place of an international Data Science and Machine Learning competition organized by IEEE. The objective of the challenge was to detect never-before seen anomalies in the behavior of a heterogeneous multi-robot system using data from different types of sensors.
• 🔧 Python, Tensorflow, Keras, OpenCV, Pandas
• 📏 LSTM + ConvLSTM Autoencoders, Frame Prediction Models, Data Fusion
• Scrapping Twitter website :
• Personal initiative to level-up my skills in Fullstack development for the Digital Tech Year program admission. TwitterScrap is a simple tool to retrieve trending thread and tweets off of Twitter and display the number of tweets per thread on a map.
• 🔧 Angular, NodeJS, Bootstap, Socket.io, REST API

Sep 2019 – Jul 2021 · 1 yr 10 mos

• Exploration and mapping of a building floor by an autonomous swarm of robots :
• One year project in which the goal was to autonomously explore and map the floor of a building using a heterogeneous swarm of small ground robots. My role along with my teammate was to develop control algorithms to efficiently explore the surface of a floor, and create a graphical simulation of the functionning of the entire swarm.
• 🔧 Python, Pygame, MPI4py
• 📏Graph Search, SLAM, Reinforcement learning
• Machine learning and computer vision projects :
• Face detection and categorization algorithm using convolutional neural networks.
• Android application to identify what types of clouds are present on an image, and infer the probability of rain in the near future.
• 🔧 Python, Tensorflow, Keras, OpenCV, Kivy
• 📏 CNN, Transfer Learning, Fine-Tuning
• Blockchain-based healthcare supply chain monitoring system :
• One-week challenge creating a blockchain web application from scratch. The objective was to use a blockchain to secure a healthcare supply chain by ensuring the traceability and authenticity of drugs.
• I was in charge of linking the Flask backend to the Angular frontend, as well as the frontend itself and some aspects of the backend.
• 🔧 Python, Angular, Flask
• 📏 Blockchain theory, elliptical curve cryptography.
• Project in partnership with ONERA:
• One month project in partnership with ONERA on the optimization of the shape of a wing to reduce the drag. Simulation of the airflow around different types of wing and entire spacecrafts.
• 🔧 Python, FEniCS Project
• 📏 Parallel and distributed computing, partial differential equations, finite element method