About Me
- Name : Maxence Leguéry
- Age :
- Occupation : Engineering student in computer science
- University : ENSTA Paris
Education
Master of science in engineering, ENSTA Paris, Palaiseau, France
Top 10 Graduate school of engineering in France
Multiple courses : Sciences, language, economics,
culture
Specialization in computer science (Pytorch, OpenCV,
Machine Learning, Deep Learning)
CPGE PTSI-PT*, Lycée Gustave Eiffel, Bordeaux, France
Advanced Physics and Mathematics Class
2 years of intense preparation for application to
graduate schools
Baccalauréat S, Lycée Fernand Daguin, Mérignac, France
French national academic qualification after secondary
education
Graduated with honor
Experience
Aug 2024
Machine learning Internship, Palaiseau, France
Deep learning for Visual Anomaly Detection.
Feb 2024
Research Internship U2IS, ENSTA Paris, Palaiseau, France
Deep learning : Uncertainty estimation and bayesian neural networks.
Aug 2023
Research Internship CPHT, École Polytechnique, Palaiseau, France
General relativity : Quasinormal modes of different
spaces-time and numerical resolution.
[Quasinormal modes in
curved spaces-time]
Jan 2023
Vice-President Technimage
Vice-President of the audiovisual club of ENSTA Paris.
Awarded best scenario with the Technimage team at Le Rush 2022
organized
at CentraleSupélec.
Internship Alsymex Laseris
Defense & Security, Energy, Aeronautics & Space and
Large Scientific Instruments.
Work in cleanroom.
Preparation for the implementation of a Computerized
Maintenance Management System (CMMS).
Volunteer organizer TFJM²
French Tournament for Young Mathematicians.
Organization of the TFJM² final at ENSTA Paris.
Professional Experiences
Engineer Internship in Deep Learning
Visionairy | February 2024 - August 2024
- Implemented AI solutions for anomaly detection in industrial environments.
- Deployed machine learning models in production using Docker and Azure.
- Collaborated with cross-functional teams to ensure smooth integration and continuous improvement of AI systems.
Research Internship in Deep Learning
U2IS Laboratory, ENSTA Paris | September 2023 - February 2024
- Conducted research on uncertainty estimation to mitigate overconfident AI model predictions.
- Collaborated with a teacher-researcher, resulting in a paper submitted to CVPR.
- Contributed to advancing deep learning techniques for robust and reliable AI models.
Skills
Programming Languages
- Python
- C++
- JavaScript
- Java
- Rust
- MatLab
Deep Learning & AI Tools
- PyTorch
- Scikit-Learn
- OpenCV
- TensorFlow/Keras
- MLFlow
Web Development
- HTML
- CSS
- JavaScript
Tools & Technologies
- Git
- Docker
- Linux
- Azure
- CI/CD (GitHub Actions)
Languages
- French (Native)
- English (Fluent)
- Spanish (Basic-Intermediate)
- Japanese (Basic)
Projects
Using C++, I developed my own realistic 3D renderer with raytracing. I used Nvidia CUDA to
massively accelerated calculations.
Features :
- Can render triangles and polygons decomposable in triangles
- Bounding volume hierarchy (BVH)
- Material : support light emission, color, glossiness, transparency
- Moving camera and adaptable parameters (fov,image size)
- SDL2 live screen with keyboard control of the camera
- Exportation in PNG format
- Support simple OBJ objects
FixMatch is a semi-supervised machine learning algorithm aiming to train a ML model using a
small amount of labelized examples by pseudolabeling many unlabelized examples to extend the
train dataset.
Kihyuk Sohn and al.
2020. FixMatch: simplifying semi-supervised learning with consistency and confidence. In
Proceedings of the 34th International Conference on Neural Information Processing Systems
(NIPS'20). Curran Associates Inc., Red Hook, NY, USA, Article 51, 596-608.
Amit Chaudhary 2020. The
Illustrated FixMatch for Semi-Supervised Learning.
3D version of the famous game : Tetris. Written in C++ using OpenGL.
Scientific Papers
Title of the Paper
Published in: Journal/Conference Name (e.g., CVPR 2024)
A brief summary of the paper's key contributions, methodology, and results. This paper investigates...
Title of the Paper
Published in: Journal/Conference Name (e.g., NeurIPS 2024)
This paper explores the use of deep learning for addressing challenges in... It presents a novel method...
