This project was made in parallel during his time as a student researcher at the AMILAB of Doshisha University. The main idea of the simulation is to use genetic algorithms to simulate what genes would survive the test of time and generations. A secondary idea is to evaluate how resource distribution and competitiveness between species would influence survival.
C++
Unreal Engine
Genetic Algorithms
Algorithm Optimization
Debugging
This project was made during my final year as a computer engineering student in Polytech Marseille. Soccer Vision App is a computer vision application based on the work of Jac99 (repo). It uses Pytorch and OpenCV to perform object detection in videos. The project aims to extend the functionalities of this application and integrate it into a web interface via an architectural framework, using Django.
Python
Pytorch
OpenCV
GitHub
Debugging
Django
Software Architecture
This project was also made during my final year as a computer engineering student in Polytech Marseille. The project is part of the study of vision-language machine learning models (LLM-VM). In order to evaluate this type of model, real images are generally used. The disadvantage of this approach is that it depends on data that cannot be controlled (e.g., bias in the data). Open-access synthetic datasets such as CLEVR already exist, but they are not realistic. The aim of this project is to semi-automatically create a dataset of annotated synthetic images, using Python and 3D modeling software, in this case Blender.
Python
Blender
JSON
Computer Vision
This project explores the use of Unreal Engine 5 and Blueprint scripting to create an immersive virtual reality experience with 4 challenges.
Unreal Engine 5
Blender
Meta Quest 2
No image available