top of page

PROJECTS

Asynchronous Sensor Fusion for Continuous-Time SLAM

Student Semester Project, ETH Zürich - 2023

For my Master's Degree at ETH, I am doing a semester project at the Vision for Robotics Lab (V4RL) of ETH.The subject is "Asynchronous Sensor Fusion for Continuous-Time SLAM", and my work is based on the HyperSLAM library developed in V4RL.

The main challenge in Simultaneous Localization and Mapping (SLAM) is that it was developed following a discrete-time approach. This approach has the advantage of being easier to implement and was very successful until now. However, when using high rate sensors like IMUs or asynchronous sensors, like Event Based Cameras, discrete-time approach does not allow to make the best use of these sensors. Furthermore, Continuous-Time approaches can lead to significant parameter reduction.

During this project:

  • I am deepening my knowledge in Computer Vision.

  • Participating to research in this captivating field.

  • Applying my knowledge using C/C++ on a Linux based system and using DevOps tools such as GitHub.

Multi-agent autonomous car's algorithm

Course Project - ETH Zürich - 2022

In the context of my Planning and Decision Making for Autonomous Robots, I developed an algorithm for autonomous driving cars tested in simulation in a multi-agent setup. The program was developed in Python and used the dg-commons library. The results from the simulations were very positive and my project was a success. I implemented well known algorithms such as RRT* or Dubins Path.

During this project:

  • I implemented the main path planning algorithms.

  • I implemented a PID for the trajectory following.

  • I used Python and GitHub for the development.

Autonomous driving

Autonomous driving

Autonomous driving
Rechercher une vidéo...
Because it cannot always run smoothly...

Because it cannot always run smoothly...

00:05
Lire la vidéo
A successful trip

A successful trip

00:22
Lire la vidéo
Another successful trip

Another successful trip

00:17
Lire la vidéo
A last successful trip

A last successful trip

00:26
Lire la vidéo

Computer Vision - Building blocks of the SLAM pipeline

Course Exercises - ETH Zürich - 2022

In the context of my Vision Algorithms for Mobile Robots, I realized each week a building block of the SLAM pipeline. The codes I developed were based on a skeleton given by the teaching assistants.

During these exercises:

  • I implemented the main vision algorithms: Harris Corner Detectors, Bundle Adjustment, SIFT Feature Detector, PnP, Lucas-Kanade Tracker...

  • I implemented these algorithms with MATLAB.

Computer Vision Projects

Computer Vision Projects

Computer Vision Projects
Rechercher une vidéo...
Harris corner detector and tracking

Harris corner detector and tracking

00:19
Lire la vidéo
Augmented reality cube

Augmented reality cube

00:24
Lire la vidéo
PnP algorithm

PnP algorithm

00:07
Lire la vidéo

Compilation of course projects

Course Projects

During my studies at ECE Paris, I had one project per course. This gave me the opportunity to always apply the concepts I learned in class.

In Programming classes I had the following projects:

  • Video game programmed in C with Alegro library.

  • SVG Landscape generator in C++.

  • C++ programs implementing the main graph theory algorithms.

  • Java school timetable and course selection application linked to a database.

  • Ricardo-like second hand website using PHP, JavaScript, HTML, CSS.

In Electronics classes I had the following projects:

  • Miniaturized smarthome.

  • Electronic attendance system with student card.

  • Microcontroller based "Snake" video game using an IMU to direct the snake.

  • Two VHDL projects to display data on a screen with VGA resolution.

Car dashboard for engine cooling and surveillance

Personal Project

As a personal project and challenge I developed a car dashboard for an old Renault 4. The goal of this project was to be able to see the engine's temperature while driving and to optimally trigger the two engine cooling fans.

To do so, I used an NTC temperature probe and a GLCD screen coupled with pushbuttons to configure the fans triggering ranges. I used an ATMega 2560 microcontroller and designed the PCB with Eagle. 

The device also allows me to read data from other probes such as gas level or oil pressure.

This device has been running successfully for 3.5 years now.

bottom of page