AmirHossein Eshghi

AmirHossein Eshghi

Master's Student in Software Engineering | Non-degree Student, Clemson University | Specializing in XAI & Computer Vision

About Me

I am a highly motivated Master of Science student in Software Engineering at the University of Birjand, currently ranked 1st in the department with a perfect 20/20 GPA. I am also advancing my knowledge as a non-degree student in Computer Science at Clemson University. My research primarily focuses on Explainable AI (XAI) in computer vision, specifically proposing novel approaches using genetic algorithms to enhance Class Activation Maps (CAMs). Furthermore, I am actively expanding my expertise into Agentic AI and NLP pipelines. I am deeply driven by the goal of bridging the gap between complex deep learning models and practical, trustworthy clinical and industrial interpretability.

Education & Transcripts

M.Sc. in Software Engineering 2024 - Present
University of Birjand, Iran
Original GPA: 20.00 / 20 (Rank: 1st in department)
Converted GPA (USA): 4.00 / 4.00 (Based on WES Standards)

Master's Thesis: I am currently researching Class Activation Maps (CAMs) in computer vision (XAI) and proposing a new approach using genetic algorithms to provide more accurate heatmaps.

Supervisors: Dr. Seyyed Ali Hosseini (University of Birjand), Dr. Hamid Saadatfar (University of Birjand), and Dr. Siavash Bigdeli (Technical University of Denmark - DTU).

Selected Graduate Coursework:

Course Name Grade (20) USA (4.0) Europe (ECTS)
Machine Learning20.004.0 (A)A (Excellent)
Data Mining20.004.0 (A)A (Excellent)
Advanced Algorithms20.004.0 (A)A (Excellent)
Distributed Cloud Computing20.004.0 (A)A (Excellent)
Non-Degree Student, Computer Science 2025 - 2026
Clemson University (Remote)

Attending online courses through the Coursera platform.

B.Sc. in Computer Engineering 2019 - 2024
University of Birjand, Iran
Original GPA: 18.66 / 20 (Rank: 1st among 125 students)
Converted GPA (USA): 3.83 / 4.00

Specialization in Artificial Intelligence.

Professional & Research Experience

Research Collaborator (Agentic AI & NLP) 2026 - Present
Supervised by Dr. Siavash Bigdeli
  • Agentic AI & Semantic Data Defragmentation: Developed a multi-agent NLP pipeline for semantic reconstruction of fragmented datasets using Large Language Models (LLMs).
AI & Data Science Researcher 2023 - 2025
Razi Hospital & Ministry of Health, Iran
  • Neurological Disease Detection (AVM): Built a Vision Transformer-based model for automated Arteriovenous Malformation (AVM) detection from MRI scans.
  • Large-Scale COVID-19 Analytics: Analyzed large-scale clinical data to extract patterns for patient care optimization.
Machine Learning Engineer (R&D) 2023 - 2025
Kavir Tire (Largest Tire Manufacturer in Iran)
  • Predictive Maintenance System: Designed a predictive maintenance model to forecast failures in curing press machines.
  • Manufacturing Optimization: Developed a deep learning model to predict and optimize steam consumption in manufacturing.
Hydrological Deep Learning Researcher 2025
Iran Meteorological Organization (IRIMO)
  • AI-Driven Flood Risk Mitigation: Developed ML/DL models for real-time river runoff prediction and flood risk forecasting.
Teaching Experience 2021 - Present
University of Birjand, Iran
Visiting Lecturer (2024 - Present)
  • Delivered lectures for undergraduate computer engineering courses.
Laboratory Teacher (2024 - Present)
  • Instructed lab sessions for Database Lab and Microprocessor Lab.
Teacher Assistant (2021 - Present)
  • Assisted with courses including: Fundamental of Programming, Data Structures, Microprocessors, Theory of Language and Automata, Computer Vision, Computational Intelligence.

Selected Publications

EvoCAM: A New Evolutionary-Based Class Activation Mapping Method
European Conference on Computer Vision (ECCV), 2026. (Under Review)
Bias, Fairness, and Equity in Medical AI (Book Chapter)
In: Trustworthy and Explainable AI for Healthcare. Springer. 2026 (Accepted, In Press)
Enhancing Runoff Prediction through Feature Engineering and Cluster-Specific Modeling
Water Resources Management JCR Q1, 2025 (Accepted)
A Multi-Timestep LSTM Ensemble Regressor for Enhanced Short-Term Runoff Prediction
Journal of Forecasting, 2025 (Under Review)
Insomnia in University Students: A Review of Etiological Factors and the Advancement of Data Analytic Techniques
Sleep and Breathing, Springer, 2024 (Published)
A Hybrid AI Framework for Academic Advising: Integrating Ensemble-Based Grade Prediction and a Rule-Based Expert System
Computers & Education, Elsevier, 2026 (Submitted)
Improving Runoff Prediction through Feature Engineering and Cluster-Based Modeling
1st National Conference on Hydro Informatics and Artificial Intelligence in Water Engineering, 2025 (Best Paper Award)
Image Retrieval Using Extracted Features and Local Binary Patterns
Informatics Conference, Tehran, 2023 (Best Paper Award)

Honors & Awards

Technical Skills

Python (TensorFlow, PyTorch, Keras) C/C++ C# SQL Scikit-learn OpenCV Git LaTeX

Languages

Persian: Native
English: Professional Working Proficiency