Tehran · Shahid Beheshti University
Computer engineering student and machine learning researcher. I Interested in Machine Learning, Computer Vision, and neuro-fuzzy systems.
Models that stay interpretable under messy data, and vision methods that recover document structure when OCR fails.
IEEE IICAI 2026 · pp. 1–6
A modular, OCR-agnostic framework for complex Persian tables. Evaluated on a dedicated real-document dataset and stronger than current OCR and VLM baselines.
Neurocomputing · 660:131866 · 2026
A neuro-fuzzy classifier for incomplete and non-separable patterns: feature selection, relaxed rule activations, and correlation-aware rules without a black box.
Manuscript in preparation
In preparation
How fuzzy BLS models are built, where they work, and what remains open for scalable interpretable fuzzy-neural systems.
B.Sc. in Computer Engineering, with research in neuro-fuzzy learning and computer vision.
B.Sc. in Computer Engineering. GPA 19.13 / 20.
Diploma in Mathematics. Final average 19.84 / 20. NODET school (National Organization for Development of Exceptional Talents).
Undergraduate research that became peer-reviewed papers.
Designed and implemented GRIFFIN, ran classification experiments, and helped write the Neurocomputing paper.
Proposed the table-detection idea, built the framework under Dr. Mohsen Ebrahimi Moghaddam, and served as first author of the IEEE IICAI paper.
Head TA and TA across core CS and math courses, plus workshops that turn lectures into working code.
Dr. Salimi-Badr. End-to-end ML workshop.
Dr. Malek. NumPy workshop.
Dr. Vahidi-Asl. Guest lecture on thread programming.
Dr. Abdoos. Five review sessions.
Guest lectures and technical workshops.
Guest lecture, Advanced Programming.
Linear Algebra workshop. Materials
Five instructor-led review sessions.
Workshop for Introduction to Machine Learning.
Industry engineering, faculty publishing, and IEEE programs.
Node.js and Express microservices and APIs.
SBU CSE Publication, with Prof. Shams Aliee.
Secretary, Professional Activities Committee. Head of IAI student committee.
Selected work in vision, architecture, and scientific NLP.
Detects a printed board, reads digits with a CNN, repairs OCR errors with Sudoku constraints, and draws the solution back onto the photo.
Pipelined MIPS processor in Verilog: datapath, core stages, and simulation through the pipeline.
Chunks PDFs, retrieves with embeddings and ChromaDB, and answers questions with OpenAI or local Llama in Streamlit.
Languages, libraries, and the ones I speak.
Based in Tehran. Happy to talk about research, teaching, or collaboration.
