I am a second-year Ph.D. student in Computer Science, supervised by Dr. Sergey Plis, at TReNDS (joint Georgia Institute of Technology/Georgia State University/Emory University Center for Translational Research in Data Science and Neuroimaging). My research interests include machine learning, computer vision, transformer and attention mechanisms, neural scene representations, and neuroimaging data science. I am particularly interested in looped transformer and attention-based architectures that iteratively refine visual representations for robust perception, 3D understanding, and neural rendering.
🔥 News
- 2026.09: 📥 One paper submitted to NeurIPS 2026 (AIM Workshop).
- 2026.07 — First author: 🎉 Our paper, “FlexSplat: Flexible Feed-Forward 3D Gaussian Splatting without Point Cloud Correspondence”, was accepted to BMVC 2026.
- 2026.07 — First author: 🎉 Our paper, “Confident but Unreliable: A Behavioral Safety Audit of Vision-Language Models on Brain MRI”, was accepted to ACM AI 2026.
- 2026.06: 📥 One paper submitted to IEEE MLSP 2026.
- 2026.06: 📝 Completed two paper reviews for IEEE MLSP 2026.
- 2026.05: 📥 One paper submitted to ACM AI 2026.
- 2026.04: 📥 One paper submitted to BMVC 2026.
- 2025.06: 🎉 Experience - Attended CVPR 2025.
- 2025.01: 🎉 Started my Ph.D. studies at Georgia State University.
- 2023.08: 🎉 Graduated from Amirkabir University of Technology (Tehran Polytechnic).
💻 Projects and Publications
FlexSplat: Flexible Feed-Forward 3D Gaussian Splatting without Point Cloud Correspondence
Accepted to BMVC 2026
Acceptance ratio: 28%
2026.07 · First author
Project: Website
Code: GitHub
arXiv: 2608.07937
Confident but Unreliable: A Behavioral Safety Audit of Vision-Language Models on Brain MRI
Accepted to ACM AI 2026
2026.07 · First author
Code: GitHub
arXiv: 2608.02790
Behavior Analysis of Cars Drivers Based-on Computer Vision
Computer-vision system for analyzing real-world driver behavior.
📖 Educations
-
2025.01 – Present, TReNDS — PhD in Computer Science
GPA: 4.0 / 4.0
Research Focus: Fixed-point self-attention and looped transformers for structured reasoning — recasting attention as a DEQ-style fixed-point iteration, analyzing convergence/gradient behavior and inference efficiency, and testing controlled PyTorch ablations on tasks such as ARC and Sudoku. -
2018.09 – 2023.07, Amirkabir University of Technology (Tehran Polytechnic) — BSc in Computer Engineering
Major: Artificial Intelligence | Minor: Computer Networks
BS Thesis: Behavior Analysis of Car Drivers Based-on Computer Vision
GPA: 3.9 / 4.0