Hello! 👋

I'm Ahmed Hassayoune

ML & Computer Vision Engineer

About Me

I am a graduate of EPITA and Sorbonne University, where I completed a double degree Master's program specializing in image processing, computer vision, and medical imaging. What drives me is the opportunity to bridge AI with real-world healthcare applications. I aspire to design and deploy solutions that make medical diagnostics faster, more reliable, and accessible to patients and practitioners worldwide.

Python Numpy PyTorch TensorFlow OpenCV C C++ CUDA Docker Medical Image Analysis

Education

Master's Degree (M2), Image Track

Sorbonne University

2024 - 2025
  • Level: Highest Honors
  • Double degree program with the IMAGE major at EPITA
  • Relevant Coursework: Computer vision, Pattern recognition, Deep learning, Advanced image processing, and Medical imaging.

Computer Engineering Degree, Image Track

EPITA

2020 - 2025
  • Top of the class in both years of the preparatory cycle
  • Exchange semester at Hanyang University in Seoul, South Korea
  • Specialization in the IMAGE major: Image Processing and Computer Vision

Abroad Exchange Semester

Hanyang University

2022 (6 months)
  • Level: GPA 87/100
  • Participated in a one-semester exchange program at Hanyang University, Seoul, as part of computer engineering studies
  • Relevant Coursework: Machine Learning, Linear Algebra, Data Structures and Object-Oriented Programming

Experience

Final-Year R&D Internship

General Electric HealthCare

Mar 2025 - Sep 2025 (6 months)
  • Developed a 3D generative model for pulmonary nodule synthesis in CT imaging
  • Designed conditioning methods to control generation along morphological and pathological aspects
  • Created synthetic CT scans and enhanced detection model performance
  • Annotated over 1,300 lung nodules (3D segmentation and labeling) to enrich training datasets
  • Collaborated with an R&D team (AI researchers, clinicians) and presented findings in internal seminars

Full-Stack Web Development Internship

DXOMARK

Sep 2023 - Feb 2024 (6 months)
  • Developed a visualization interface for smartphone camera scoring used by engineers and product teams
  • Migrated a large-scale app from Vue 2 to Vue 3 (20+ dependencies)
  • Designed and implemented backend services, integrated into an existing modular architecture
  • Contributed in an Agile Scrum team with developers, and image QA engineers, improving release cycles

Projects

01

Alignment of Mass Spectrometry Data

In collaboration with Foch Hospital, Paris-Saclay University, and Versailles Saint-Quentin University, I developed fast and robust alignment methods for data from 2D gas chromatography coupled with mass spectrometry (GCxGC-MS). This technology is used in clinical research to analyze volatile organic compounds (VOCs) in exhaled breath, enabling non-invasive diagnosis of lung diseases such as asthma, cancer, and COVID-19.

Image ProcessingPattern RecognitionComputational optimizationBioinformaticsPythonMatlab
02

3D Brain Tumor Analysis

This group project focused on tracking and analyzing tumor changes in medical images. Using registration and segmentation techniques, we aligned and processed scans to highlight differences over time, then built visualizations to make these changes easier to interpret. We compared several algorithms through benchmarking to determine the most effective methods, and implemented the solution in Python with ITK and VTK for image processing and 3D visualization.

PythonITK/VTKImage RegistrationImage Segmentation3D Visualization
03

Implementation of SIFT Algorithm

Implemented from scratch the SIFT algorithm in C++ for robust feature detection and matching in images. Optimized the implementation for real-time processing, enabling efficient and rapid feature detection and matching. Applied to automatic stitching of fragmented images for scene reconstruction.

C++Computer VisionFeature Detection and MatchingReal-Time Optimization
04

Real-Time Video Background Subtraction GPU Plugin

GStreamer plugin capable of separating moving objects from the background in real-time video streams. The plugin leverages GPU-accelerated image processing to achieve high performance. The approach involves analyzing each video frame against a dynamic background model, detecting motion by identifying differences, filtering noise, and generating a clean foreground mask. The background model is continuously updated to adapt to changes in the scene, ensuring robust and efficient object detection.

CUDAC++Image ProcessingComputer Vision
05

Adaptive Quad Mesh Simplification in Python

This project implements a Python tool using the Blender API to simplify 3D quad meshes while keeping as much detail as possible. By reducing mesh complexity efficiently, it makes 3D models lighter and easier to use in graphics or simulation, following techniques from a 2010 research paper. The implementation was done completely from scratch, including mesh triangulation, Tri-to-Quad conversion, and adaptive simplification.

PythonBlenderBlender Python APIComputer GraphicsGeometric optimization