Conducted comprehensive literature reviews to gather relevant information for new research ideas and projects.
Analyzed and interpreted large scale image dataset (Over 4 millions) to draw meaningful conclusions and build roadmaps for object detection , classification and lane detection models.
Optimized image augmentation techniques to generate more balanced and robust dataset ,reducing our models FP rate by 80%.
Trained multiple computer vision based models achieving up to 99.6% of accuracy, validated them using field experiments and deployed them to Windows GPU machines and Cloud pipelines for Live pilot purposes.
Computer Vision Specialist
AI For Women Empowerment -Hackathon
10.2022 - 12.2022
The team came up with a Mobile App to help Morocco's illiterate people hear and understand any paper in Arabic,their native language .
Supported the design of the idea , planned and executed data collection and Image to text models fine tuning .
The App won a first place at AI For Women Empowerment Hackathon and the solution was purchased by UM6P university.
Computer vision researcher
IAV Institute
02.2020 - 10.2023
Led an End-to-end object detection project applied to automated aircraft maintenance inspection
Augmented the initial database by 98 % due to the lack of aircrafts dents images .
Determined and evaluated the best model for my use case within 3 of state-of-the-art architectures : U-NET, Attention U-NET and TransU-Net as ViT.
Achieved a high performance with 84 % of recall with big optimization of time of dents detection from 2 hours to 20min per aircraft.
Presented the project outcomes at IJCAI 2021 workshop.
Engineer Intern
ETAFAT
05.2020 - 07.2023
Designed an automation workflow for Volume calculation using 3D Cloud point data collected via USV.
Tested the procedure in a real bathymetric project and optimized the time of calculation by 96 %.
Edited a complete guidebook to integrate the new workflow in production pipelines.