Zeinab Rahbar

**AI Researcher

Graph Learning & Geometric Deep Learning**

📍 Isfahan, Iran	📧 ZeinabRahbar2022@gmail.com	📞 +98 9100724610
🔗 LinkedIn	💻 GitHub

Research Summary

AI researcher with a strong background in graph neural networks, latent graph learning, and multimodal representation learning, with a growing focus on theoretical and efficient processing of graph-structured data. Experienced in learning graph structures from data and analyzing information propagation and generalization behavior in graph-based models. Interested in graph signal processing, sampling and reconstruction of graphs, graph compression, biomedical imaging and theory-driven geometric deep learning.

Education

M.Sc. in Computer Engineering, Artificial Intelligence | University of Isfahan (2022–2025)

Thesis: Developing graph neural networks for multimodal non-graph textual and visual data
Grade: 18.60/20
Supervisors: Dr. Peyman Adibi, Dr. Alireza Darvishy (ZHAW, Switzerland)

B.Sc. in Computer Engineering

Arak University (2018–2022)

Selected Research & Projects

Latent Graph Learning for Multimodal Disease Prediction – Model that infers latent graph topologies from heterogeneous data (MRI, fMRI, clinical tests). Evaluated on TADPOLE and ABIDE datasets.
GNN-based Visual Question Answering (VQA) – Modeled visual/textual entities as graphs using GNNs for propagation. Experiments on VQA v2 dataset.
Graph Convolutional Networks with random weights – Implemented approach from paper (no public code available).

Professional Experience

Machine Vision Engineer | HoopadVision, Isfahan (Feb 2024 – Mar 2025)

Improved real-time fire detection accuracy with time-series analysis
Built human action recognition system (running, falling) – 4 camera streams @ 15 FPS on single GPU

Teaching & Mentorship

Teaching Assistant: Machine Learning, Computer Basics, HCI, Discrete Mathematics – University of Isfahan & Arak University
Python Instructor – Yasan Academy (2023–2024)

Technical Skills

Research Interests: GNNs, Geometric Deep Learning, Multimodal ML, Biomedical Data, Medical Image Processing

Languages: Python, C++

Frameworks: PyTorch, TensorFlow, PyTorch Geometric, DGL, Ultralytics, FastAPI

Tools: Docker, Git, Linux, VS Code

Languages

English: IELTS 7.0 (L:7.5, R:7, W:6.5, S:7.5)
Persian: Native
French: Beginner (A1)

Open to research collaborations and PhD opportunities.