Marc Szafraniec
Research Engineer at Facebook AI
France
Summary
🤩
Rockstar🎓
Top SchoolMarc Szafraniec is a research engineer based in France with a decade of experience applying machine learning to vision and finance problems, currently advancing computer vision at Facebook AI. Trained at École Polytechnique and ENS Cachan with a Master’s in Data Science and Vision & Machine Learning, he blends strong mathematical foundations with practical ML engineering. His contributions to Detectron2 include implementing DensePose components and test-time augmentation, showing hands-on impact in a widely used open-source detection framework. Past roles span quantitative research in finance and high-accuracy extraction of tabular financial data from emails, reflecting a knack for turning research into robust, production-ready systems. Outside engineering he pursues filmmaking, bringing a creative perspective to technical problem solving.
11 years of coding experience
3 years of employment as a software developer
Baccalauréat with Honors, Major in Sciences, Baccalauréat with Honors, Major in Sciences at Lycée Stanislas, CannesMaster of Science (M.S.), Data Science, Master of Science (M.S.), Data Science at École PolytechniqueSecond Year of Classe préparatoire aux Grandes Ecoles, Mathematics, Physics, Computer Sciences, #43 at the Polytechnique entrance exam, Second Year of Classe préparatoire aux Grandes Ecoles, Mathematics, Physics, Computer Sciences, #43 at the Polytechnique entrance exam at Lycée Masséna, NiceMaster 2 (M2), Mathematics, Vision and Machine Learning, Master 2 (M2), Mathematics, Vision and Machine Learning at École Normale Supérieure de Cachan
French, English, German, Russian, Polish
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (10)
computer-vision10
pytorch10
machine-learning10
eval10
deep-learning10
detectron10
python10
evaluation10
model-testing10
data-augmentation9
Programming languages (4)
Github contributions (5)
facebookresearch/detectron2
Jan 2020 - May 2021
Detectron2 is a platform for object detection, segmentation and other visual recognition tasks.
Role in this project:
Back-end & ML Engineer
Back-end & ML EngineerContributions:98 commits, 2 comments in 1 year 4 months
Contributions summary:Marc contributed to the implementation of the DensePose model within the Detectron2 framework. They added and modified components such as the DensePose DeepLab head and Decoder, and incorporated CSE (coarse segmentation estimation) features, including the development of texture transfer visualizers. The contributions involve changes to configuration files, integration of new datasets, and optimization of the evaluation pipeline. Furthermore, the user made changes to integrate the test-time augmentation functionality of the framework.
detectronobject-detectionsegmentationvisual-recognition
MarcSzafraniec/MarcSzafraniec.github.io
Jan 2017 - Oct 2017
Contributions:73 pushes, 1 branch in 9 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.