Youssef El Assfouri

• Led sub-pixel metrology campaigns using vision for aeronautical flight tests, providing high-precision measurements of structural deformations and kinematics, used for design validation and certification, while coordinating up to 3 engineers.

• Developed custom Python pipelines in 2D/3D geometric computer vision and AI for in-flight analysis of structures and kinematics: pose estimation, 3D reconstruction, detection, identification, tracking, and sub-pixel keypoint registration, using data from 3D scanners, lidars, event-based cameras, infrared, and conventional cameras.

• Supported multidisciplinary teams (structures, flight testing, R&D) in defining vision system requirements and implementing appropriate measurement solutions.

• Contributed to a proprietary Airbus computer vision library in Python, developing an optimized module for decoding coded circular targets.

• Calibrated mono, stereo, and infrared camera systems for high-precision measurement applications.

• Evaluated and integrated new computer vision tools, SLAM algorithms, and Edge AI solutions to improve project and measurement process efficiency.

Acoustic Signal & Image Processing for Aerospace Applications

▪ Designed and implemented an event detection algorithm for low signal-to-noise ratio (SNR) aircraft background noise recordings, enabling automated identification of acoustic anomalies.

▪ Documented results and provided technical reports to stakeholders.

▪ Developed a real-time Python video game controlled by hand gestures, achieving 30.7 FPS of hand detection without GPU acceleration.

▪ Implemented a hand tracking pipeline using a pre-trained convolutional neural network (CNN) based on the SSD architecture via Google MediaPipe.

▪ Designed the computer vision and deep learning inference pipeline for gesture classification and real-time game control.