About me

mypic

I am a PhD student at School of Engineering at École Polytechnique Fédérale de Lausanne (EPFL) and a Research Assistant at Biometric Group in Idiap Research Institute under supervision of Prof. Marcel and Prof. Alahi. Recently, I was a Research Scientist Intern at J&J Innovative Medicine. Previously I got my masters in Electrical Engineering from Sharif University of Technology, under supervision of Dr. Arash Amini.

Before starting my PhD I was lead research engineer at MCI.

My main research interest is Synthetic Data Generation and its application in various Computer Vision tasks. Specifically my current approach is to study if it is possible to use the Generative Models like StyleGANs, Diffusion and Flows to further enhance the Discriminative models (e.g., Classifiers). Conceptually this have some common grounds with Analysis by Synthesis idea. In essence my work is to study that can the Visual Generative Models be USEFUL beside the entertainment industry or not.

ScoreMix: Synthetic Data Generation by Score Composition in Diffusion Models Improves Recognition

ScoreMix Presented at ICML 2026

Presentation of ScoreMix: Synthetic Data Generation by Score Composition in Diffusion Models Improves Recognition at the International Conference on Machine Learning (ICML 2026) in Seoul, South Korea.

July 20, 2026 · 1 min · Parsa Rahimi
J&J Innovative Medicine

Research Scientist Intern at J&J Innovative Medicine

Joined Johnson & Johnson Innovative Medicine as a Research Scientist Intern focusing on biomedical AI and data science research.

June 1, 2026 · 1 min · Parsa Rahimi
AugGen: Synthetic Augmentation Can Improve Discriminative Models

AugGen Accepted at NeurIPS 2025

Paper on AugGen: Synthetic Augmentation Can Improve Discriminative Models accepted at the Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS 2025).

September 19, 2025 · 1 min · Parsa Rahimi
Synthetic to Authentic: Transferring Realism to 3D Face Renderings for Boosting Face Recognition

Oral Presentation at ECCV Workshop 2024

Paper on Synthetic to Authentic: Transferring Realism to 3D Face Renderings for Boosting Face Recognition accepted for oral presentation at the ECCV 2024 Workshop on Synthetic Data in Computer Vision.

September 1, 2024 · 1 min · Parsa Rahimi

Paper Presentation at ICASSP 2024

Paper introducing Deep Variational Privacy Funnel (DVPF) accepted at the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2024).

February 11, 2024 · 1 min · Parsa Rahimi