First year PhD student at Inria
Studying Gaussian splatting and LiDAR for SLAM
Projects
You can have a look at my projects on this page
ESA - Master thesis
My master's thesis, conducted at the European Space Agency (ESA), focuses on the scaled-down model of the ExoMars rover, which is equipped with advanced technologies such as a long-range WiFi antenna, a LiDAR system, GNSS and RTK modules, and a gyroscope etc. within the ROS2 middleware.
The primary objective of my research was to perform field tests to collect crucial data, including RGB images, depth measurements, and force-torque data on the wheels.
These datasets were meticulously extracted, processed, and utilized to train a neural network designed to predict the traversability of a given patch of land from the rover's perspective. This predictive capability is vital for ensuring that the rover can navigate safely and efficiently over various terrains.
The work involving the network training is still ongoing.
This research plays a significant role in preparing for future Mars missions by providing tools and methodologies to ensure the safety and effectiveness of exploration rovers.
​
Link to the field test paper
​
IRDP - APRIL
APRIL is a research project developed within the framework of the interdisciplinary research discovery project in partnership with inria Nancy grand Est (Larsen Team). The goal of this project is to design a gripper robot able to detect, recognize and grasp objects at range. This project is based on a programmable robotic "Bioloid" kit from Robotis.
​
- Custom Yolo v4 tiny object detector
- 3D printed PLA base
- 3D printed flexible PLA hands with voronoï pattern for adaptable grasping and deformation
NR-DS NeRF
Completed my internship of second year of engineering school (M1) within the ImViA lab supervised by the Vision for robotic team (ViBot) where I worked on the Neural Radiance Fields (NeRF) thematic. I especially developped a depth supervised, non-rigid NeRF (NR-DS-NeRF).
​
- Studied NeRFs rigid and non algorithms
- Add of Depth supervision on pytorch NR-NeRF algorithm
- Creation of a Non rigid dataset for NR-DS NeRF evaluation
Urbanloop obstacle avoidance (on going)
This project aims to secure the tracks of the autonomous Urbanloop. On the one hand, by integrating a computer vision system allowing to detect movement and what is on the rails at the level of the platform when it arrives in the station. On the other hand, by embedding a vision system inside the capsule to monitor the capsule's surroundings to avoid collisions.
​
- Background substraction with Gaussian blur
- Yolo algorithm for object detection after movement detection
- Hough transform for rails detection
Photo by : Urbanloop
Photo by : https://www.timothee-leblond.com/artseio
Ant-assisted design (on going)
This project consists in the observation and the study of the behavior of a population of ants in case of distress in order to model their behavior. At the crossroads between biology and engineering, the interest of this new study is to put forward new design proposals that can improve the layout of public spaces or buildings where the number of people can be a danger in case of evacuation.
Bio-informatics towards reasearch (On going)
Development of Bio-info algorithms to identify recombination sites in genomes.
​
- Genome analysis
- Panda & matplotlib librairies
- Datascience
Photo by : https://www.timothee-leblond.com/artseio
ImageJ object detection
This project aims to create ImageJ macros and use image processing techniques such as otsu thresholding and object property calculation to detect and label objects in the image.
​
- Otsu thresholding
- Morphological image processing
- Object labellisation with object properties
Past projects
3 years old or more projects
Self-training OpenCV project
This project aims to deepen my knowledge and skills on the OpenCV library. From annotation, transformation and image processing to face detection by haar cascade.
​
- OpenCV
- Haar cascade
- Image Processing
Self driving CAR
This project is part of the interdisciplinary baccalaureate project I developed in my last year of high school. It consists in a car (arduino) driven by a vision system embedded on raspberry pi. The car follows a colored road and is able to recognize stop signs.
​
- Arduino based car
- Raspberry pi & camera for road and sign detection
Photo by : https://www.youtube.com/watch?v=bixaJBG5RKg