Eden Belouadah

Eden Belouadah Eden Belouadah

ML Research Scientist

Datakalab

About me

Hi! I am Eden Belouadah, an ML Research Scientist at Datakalab. I have 5 years of experience working on continual learning for image classification and object detection. At Datakalab, I am part of the context adaptation team. Our goal is to compress large object detection models and deploy them on the edge. I previously worked on fall detection and continuous optimization. Currently, I am also working on Large Language Models (LLMs) and Stable Diffusions. P.S. The website is under construction.

Interests
  • Continual Learning
  • Object Detection
  • Generative AI
  • Stable Diffusion
  • Large Language Models
  • Fall Detection
Education
  • PhD in Artificial Intelligence, 2021

    IMT Atlantique, France

  • MSc in Machine Learning, 2018

    Paris-Saclay University, France

  • MSc in Artificial Intelligence, 2017

    USTHB, Algeria

  • BSc in Maths and Computer Science, 2015

    USTHB, Algeria

Experience

 
 
 
 
 
Datakalab
Deep Learning Researcher
December 2021 – Present Paris
  • Use compression techniques on object detection models and deploy them on the edge
  • Propose continual learning methods for context adaptation without catastrophic forgetting
  • Work on large language models and diffusion models
  • Write CD/CI pipelines, unit tests, and master VSCode debugging tool
  • Deliver on-demand projects to clients
  • Write a scientific publication to validate academic results
 
 
 
 
 
CEA-LIST & IMT Atlantique
Researcher Ph.D. Student
December 2018 – December 2021 Paris
  • Propose state-of-the-art Continual Learning solutions for image classification
  • Write several scientific publications to validate the proposed methods
  • Give Deep Learning classes to Master’s students (Centrale Supélec School)
  • Give Operating system classes to Bachelor’s students (Paris-Saclay University)
  • Supervise Interns and collaborate with other Ph.D. students

Recent Publications

Quickly discover relevant content by filtering publications.
(2023). PlaStIL: Plastic and Stable Exemplar-Free Class-Incremental Learning. Proceedings of The 2nd Conference on Lifelong Learning Agents.

Cite

(2023). MultIOD: Rehearsal-free Multihead Incremental Object Detector. arXiv preprint arXiv:2309.05334.

Cite

(2021). A comparative study of calibration methods for imbalanced class incremental learning. Multimedia Tools and Applications.

Cite

(2020). Active Class Incremental Learning for Imbalanced Datasets. Computer Vision - ECCV 2020 Workshops - Glasgow, UK, August 23-28, 2020, Proceedings, Part VI.

Cite

(2018). DeeSIL: Deep-Shallow Incremental Learning. TaskCV Workshop @ ECCV 2018..

Cite

Recent Talks

Continual Learning for Image Classification and Object Detection
Continual Learning for Image Classification and Object Detection
Large-Scale Deep Class-Incremental Learning
Large-Scale Deep Class-Incremental Learning