YOUSSEF ADARRAB
![[WIP] microJEPA — JEPA From Scratch in Pure Python](/images/microjepa-brain.png)
2024 — present
Software Engineer — Backend, Ops & ML
Backend infrastructure and operational systems.
2023 — 2024
ML Engineer / Backend Engineer
ML pipelines and backend systems.
2022 — 2023
Research Intern — Deep Learning
Generative models and synthetic data.
Monk AI · Paris
2023
Tech Volunteer
Nt3awnou · Morocco
Humanitarian relief tools after the Morocco earthquake.
2023
Guest Lecturer
X-HEC Masters · Paris
Applied machine learning.
2021 — 2022
Research Assistant
SONDRA Lab · CentraleSupélec
Signal processing and machine learning research.
2020 — 2021
ML / MLOps Engineer
N+ONE Datacenters · Paris
Machine learning and MLOps infrastructure.
2021 — 2022
CentraleSupélec
Paris-Saclay
Applied Mathematics, Computer Science, AI.
2018 — 2021
EIGSI
Casablanca
Engineering.
No Village Left Behind: A Moroccan Data-driven Platform for Effective Aid Coordination
Following the catastrophic earthquake that hit Morocco in September 2023, our platform emerged to optimize relief coordination, efficiently orchestrating resources to aid those in need. This paper presents the various techniques used to collect and process requests and interventions into a clean and actionable dataset, enabling authorities and fellow NGOs to efficiently extend aid to the affected areas.
Deep Learning for the Super Resolution of SAR Images
In this study, we are interested in Synthetic Aperture Radar (SAR) data, in particular those captured during campaigns carried out by unmanned aircraft. SAR data is captured by a radar antenna placed underneath either an aircraft or a satellite and pointing to the sides. Our work considered horizontally polarised transmitting and measuring antenna (HH). From these measurements, one can detect ground movements, objects or buildings, segregate land use, etc. In contrast to optical measurements, SAR data can be captured day and night and can penetrate cloud layers. However, the quality of object detection or segmentation is dependent on the frequency of the chirp and this study investigates the ability to infer high resolution SAR images from low resolution ones.