OCR pipeline for document understanding
An editable sample project for extracting structured text from visually complex documents.
AI / ML engineer · computer vision
Samy Aouabed
I build machine learning systems centered on visual understanding, perception and intelligent systems.
Selected work
All projectsPose estimation for human action understanding
An editable sample system for reasoning about human motion from pose sequences.
Retrieval-augmented technical assistant
An editable sample RAG workflow for grounding answers in a controlled document collection.
Research notes
Research libraryDINOv2: Learning Robust Visual Features without Supervision
Editable reading note on general-purpose visual features learned without task-specific supervision.
Vision transformers for action recognition [reading note]
Placeholder note for comparing spatial and temporal attention in action-recognition systems.
Multimodal retrieval and RAG [reading note]
Placeholder note for studying retrieval across text and visual representations.
Experience
Full experienceL'Oréal Research & Innovation
[Add role]
Djezzy
[Add role]
CDTA
[Add role]
About
I work at the intersection of visual perception, machine learning systems, and applied research. This editable introduction stays intentionally factual and concise.