cv

Master of Science (Thesis)

McGill University

Electrical and Computer Engineering

Bachelor of Science

Sharif University of Technology

Computer Science

Mila EDI Scholarship - Women in AI

MILA (Quebec AI Institute)

Award valued at CAD$10,000

Graduate Excellence Fellowship

McGill University

Award valued at CAD$4,500

Iranian Nationwide University Entrance Exam

Iranian Ministry of Science

Ranked 229th among more than 140,000 participants (Top 0.17%)

Language-Guided Trajectory Traversal in Disentangled Stable Diffusion Latent Space for Factorized Medical Image Generation

CVPR (2025) MIV Workshop Proceedings

Pixel Perfect MegaMed: A Megapixel-Scale Vision-Language Foundation Model for Generating High Resolution Medical Images

MICCAI (2025) DGM4MICCAI Workshop Proceedings

Pixels Under Pressure: Exploring Fine-Tuning Paradigms for Foundation Models in High-Resolution Medical Imaging

MICCAI (2025) ELAMI Workshop Proceedings

RL4Med-DDPO: Reinforcement Learning for Controlled Guidance Towards Diverse Medical Image Generation using Vision-Language Foundation Models

MICCAI 2025

Towards Reliable Human Pose Forecasting with Uncertainty

IEEE Robotics and Automation Letters

Graduate Student Researcher

Mila - Quebec Artificial Intelligence Institute

Working on generative modeling and explainability and their application in healthcare.

• Developing fine-tuning methods for Vision Language Foundation Models.
• Exploring latent subspaces in Stable Diffusion for controlled image generation.
• [Paper submitted to MIV-CVPR Workshop](https://arxiv.org/abs/2503.23623)

Software Developer

NodeEffect LTD

Worked on blockchain technologies and led peer-to-peer cryptocurrency exchange software development.

• Developed an open-source commit review tool
• Contributed to FSharpLint, commitlint, lnp2pBot, fsx

Research Intern

University of California Irvine (UCI)

Joint research with Sharif University on food image and recipe analysis.

• Developed a hierarchical visual feature extractor for predicting ingredients and categories.
• Built a Graph-LSTM model for classifying food using recipe graphs.

Research Intern

EPFL - VITA Lab

Improved state-of-the-art models for Human Pose Prediction.

• Integrated predictive uncertainty into models (+2–5% improvement).
• Maintained deep learning library with team of 5.

Persian Spell Correction

Designed and implemented a transformer-based model for correcting Persian text spelling.

Posepred

Developed a library for integrating and comparing state-of-the-art human pose prediction models.