BAFreelancer profile translated to English.
Back to original languageAbout Bouzid
Data Scientist & Teacher in Signal Processing — Expert in Deep Learning, Computer Vision & Neural Compression
Doctor of Computer Science and postdoctoral researcher, I design and optimize advanced AI models for computer vision, image processing, and learning-based compression. I master the entire chain, from scientific prototyping to industrialization, leveraging cutting-edge architectures (Transformers, SSM, VAE, and SNN) and solid expertise in signal processing.
I am particularly involved in:
- Fine-tuning and adaptation of foundation models (Grounding DINO, Florence),
- Distillation, compression, and optimization of models for constrained environments,
- Design and deployment of AI pipelines with good MLOps practices,
- Analysis of visual data and time series,
- Support & skill transfer (teaching, popularization of technical concepts).
My approach is based on a dual strength: scientific rigor and operational efficiency. I excel at transforming complex problems into reliable, high-performing, and maintainable AI solutions, adapted to the concrete needs of both industry and R&D.
French
Native or bilingual
English
Fluent
Arabic
Native or bilingual
kabyle
Native or bilingual
Can work on-site
Paris (up to 50km)
Experience
- Université Sorbonne Paris NordPostdoctoral ResearcherRESEARCHJanuary 2025 - Today (1 year and 5 months)Paris, France• Few-Shot Object Detection in aerial images.• Knowledge distillation for model optimization, improving efficiency and performance in deep learning architectures.• Fine-tuning foundational models such as Grounding DINO and FLORENCE, refining their capabilities for specific applications.teaching signal/image processing, C language
- Laboratoire de Traitement et Transport de l'Information L2TI (Institut Galil´ ee)Doctoral Researcher in Deep Learning-Based Image CompressionSeptember 2021 - December 2024 (3 years and 3 months)Nord, France• Conducted advanced research on end-to-end deep learning models for image compression, focusing on optimizing storage and bandwidth constraints for high-resolution images• Developed three innovative approaches: – Swin Transformer without Positional Encoding (SwinNPE)
- Laboratoire de Traitement et Transport de l'Information L2TI (Institut Galilée)Research InternRESEARCHNovember 2020 - March 2021 (4 months)Paris, FranceStudy and evaluation of various motion estimation models: FlowNet, SpyNet, PWC-Net, RAFT, ARFlow,
Recommendations
Be the first to recommend Bouzid
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Ph.D. in Computer ScienceSorbonne Paris Nord University2024Ph.D. in Computer Science
- M.Sc. in Computer Science, Vision and Intelligent Machines (VMI)University of Paris (formerly Paris-Descartes)2021M.Sc. in Computer Science, Vision and Intelligent Machines (VMI)