Research

AI, machine learning, high-performance computing, and data analytics, organized by research focus.

15 projects · 20 publications & graduate works

Computer Vision

Generating, recognizing, and interpreting images, with applications in OCR and engineering inspection.

Wind-turbine inspection

About the project →

RGB and thermal computer vision for identifying defects in wind-turbine blades.

Thermal imaging · Data fusion · Defect detection

Related publications (6)

2025 · Journal article · Project publication

Enhanced Non-Destructive Testing of Small Wind Turbine Blades Using Infrared Thermography (opens the publication record in a new tab)

Infrared thermography · Defect localization · YOLO

M. Memari, M. Shekaramiz, M. A. S. Masoum, A. C. Seibi · Machines, 13(2), 108

Studies thermal-image defect detection and localization on small wind-turbine blades, including YOLOv8 and YOLOv9 methods. The controlled experiments support further investigation toward drone-based inspection.

2024 · Journal article · Related research

Deep Learning for Indoor Pedestal Fan Blade Inspection: Utilizing Low-Cost Autonomous Drones in an Educational Setting (opens the publication record in a new tab)

Drone imaging · Defect detection · Engineering education

A. A. Rodriguez, M. Davis, J. Zander, E. Nazario Dejesus, M. Shekaramiz, M. Memari, M. A. S. Masoum · Drones, 8(7), 298

Uses small drones and pedestal fans as a low-cost, controlled surrogate for blade inspection. Students apply computer vision and machine learning to simulated defects through a hands-on engineering project.

2024 · Journal article · Project publication

Review on the Advancements in Wind Turbine Blade Inspection: Integrating Drone and Deep Learning Technologies for Enhanced Defect Detection (opens the publication record in a new tab)

Drone inspection · Deep learning · Structural integrity

M. Memari, P. Shakya, M. Shekaramiz, A. C. Seibi, M. A. S. Masoum · IEEE Access, 12, 33236–33282

Reviews drone inspection, image processing, deep learning, and structural-integrity approaches for wind-turbine blades. It connects sensing and automated defect detection with practical blade-maintenance challenges.

2024 · Journal article · Project publication

Data Fusion and Ensemble Learning for Advanced Anomaly Detection Using Multi-Spectral RGB and Thermal Imaging of Small Wind Turbine Blades (opens the publication record in a new tab)

Multispectral data fusion · Ensemble learning · Anomaly detection

M. Memari, M. Shekaramiz, M. A. S. Masoum, A. C. Seibi · Energies, 17(3), 673

Combines RGB and thermal imaging with deep-learning classifiers and ensemble methods for small-blade anomaly detection. It compares models and image-fusion approaches using controlled experimental data.

2024 · Journal article · Project publication

Identification and Localization of Wind Turbine Blade Faults Using Deep Learning (opens the publication record in a new tab)

Object detection · Instance segmentation · Blade fault localization

M. Davis, E. Nazario Dejesus, M. Shekaramiz, J. Zander, M. Memari · Applied Sciences, 14(14), 6319

Compares YOLO and Mask R-CNN architectures for locating cracks, holes, and erosion in small-scale wind-turbine blades. The work extends image classification toward identifying where defects occur.

2024 · Conference paper · Project publication

Wind Turbine Blade Fault Detection via Thermal Imaging Using Deep Learning (opens the publication record in a new tab)

Thermal imaging · Deep learning · Fault detection

B. Collier, M. Memari, M. Shekaramiz, M. A. S. Masoum, A. Seibi · IETC 2024 · IEEE, pp. 23–28

Applies deep-learning methods to thermal imagery for wind-turbine blade fault detection. This conference contribution forms part of the broader imaging and inspection research developed in the related journal studies.

Synthetic image evaluation for OCR

About the project →

Evaluate generated images and use synthetic data in handwritten-digit recognition.

Generative modeling · Synthetic data · OCR

Related publications (3)

2024 · Preprint · Project publication

Advancing Generative Model Evaluation: A Novel Algorithm for Realistic Image Synthesis and Comparison in OCR System (opens the publication record in a new tab)

Generative model evaluation · Synthetic images · OCR

M. Memari, K. R. Ahmed, S. Rahimi, N. A. Golilarz · arXiv:2402.17204

Introduces a procedure for comparing synthetic images, focusing on Arabic handwritten digits and OCR. It refines image-quality evaluation to support generative-model comparison and improvement.

2023 · Conference paper · Project publication

Generative Data Augmentation for Arabic Handwritten Digit Recognition Boosting Real-time OCR Capabilities (opens the publication record in a new tab)

Generative data augmentation · Handwritten digit recognition · OCR

M. Memari, K. R. Ahmed, S. Rahimi · AIPR 2023 · ACM, pp. 312–323

Investigates generative data augmentation for Arabic handwritten-digit recognition and real-time OCR. It connects synthetic-image generation with recognition and image-quality evaluation.

2023 · Dissertation · Project publication

A New Approach to Synthetic Image Evaluation (opens the publication record in a new tab)

Variational autoencoders · Generative adversarial networks · Synthetic image evaluation

M. Memari · Ph.D. dissertation · Southern Illinois University Carbondale

Investigates conditional VAE and GAN image generation for Arabic handwritten-digit recognition. It develops the Synthetic Image Evaluation Procedure to connect image quality, feature extraction, and OCR experiments.

Chest X-ray image classification Unpublished Ph.D. research

About the project →

Unpublished Ph.D. research exploring computer vision for COVID-19 detection from chest X-ray images.

Medical imaging · Image classification · COVID-19

GridEye Historical concept

About the project →

An earlier research concept exploring aerial AI and power-quality information for grid inspection.

Aerial imaging · Sensor fusion · Infrastructure inspection

Other publications in this area (1)

2023 · Journal article

Advances in Computer Vision and Image Processing for Pattern Recognition: A Comprehensive Review (opens the publication record in a new tab)

Computer vision · Image processing · Pattern recognition

M. Memari · International Journal of Engineering and Applied Sciences, 11(05), 2896–2901

Link opens the author manuscript on SSRN.

Surveys methods and applications across computer vision, image processing, and pattern recognition. It provides a broad methodological overview rather than a single application study.

Education

Using AI to support teaching, learning, curriculum design, and human-led assessment.

Building AI Solutions

About the project →

Build useful AI solutions through interactive simulations, moving figures and an AI Learning Companion. Open to any discipline, with no prerequisites or coding required.

AI literacy · Project-based learning · AI evaluation

Related publications (2)

2026 · Accepted journal article · Related research

The Capability Ladder: Aligning AI Delegation with Human Supervision in Computing Education (opens the publication record in a new tab)

AI delegation · Human supervision · Computing education

M. Memari, G. Rudolph, K. Ruggles · Frontiers in Education · Digital Education

Accepted 20 October 2026; final formatted version forthcoming.

Proposes a framework connecting AI delegation with observable evidence of human supervision. It provides a classification procedure, supervision record, and analytic rubric; its constructed examples do not establish measured learning gains.

2025 · Conference paper · Related research

Enhancing STEM Education through Project-Based Learning: Leveraging AI Tools in Undergraduate Courses (opens the publication record in a new tab)

Project-based learning · Object detection · Automated annotation

M. Memari, H. Petit, N. Olson, F. Ahmadi · HUIC STEM/STEAM & Education 2025

Describes a project-based undergraduate capstone using YOLOv12 and automated labeling for road-hazard detection. It examines AI tools as support for data preparation, experimentation, interdisciplinary collaboration, and reflection.

Practice classroom interactions with AI student personas.

LLMs · Multi-agent systems · Classroom simulation

Related publications (2)

2026 · Conference paper · Project publication

AI-ClassSims: Multi-Agent LLM Classroom Simulation with Real-Time Feedback for Teacher Training (opens the publication record in a new tab)

Multi-agent systems · Large language models · Classroom simulation

M. Memari, K. Ruggles · IETC 2026 · IEEE

Presents a classroom simulator with multiple student agents, persona memory, group discussion, and instructional feedback. Initial evaluation examines simulated behavior and language rather than establishing teacher-training efficacy through a controlled study.

2025 · Preprint · Related research

Artificial Intelligence in Elementary STEM Education: A Systematic Review of Current Applications and Future Challenges (opens the publication record in a new tab)

Systematic review · Elementary STEM · AI-assisted learning

M. Memari, K. Ruggles · arXiv:2511.00105

Synthesizes AI applications in elementary STEM, including tutoring, learning analytics, assessment, computer vision, and robotics. It identifies questions around developmental suitability, privacy, equity, and teacher-centered implementation.

AI-assisted student-teaching evaluation aligned with USBE rubrics.

LLMs · Human-in-the-loop AI · Rubric-based assessment

Related publications (1)

2026 · Conference paper · Project publication

Accelerating Feedback, Preserving Expertise: AI-Assisted Evaluation in Teacher Preparation (opens the publication record in a new tab)

Human-in-the-loop AI · Rubric-based assessment · Teacher preparation

K. Ruggles, M. Memari · SITE 2026 · AACE / LearnTechLib

Introduces AI-STER, a human-in-the-loop approach combining language models with standards-aligned rubrics for student-teacher assessment. It focuses on supporting feedback while preserving educators’ evaluative judgment.

AI Concepts Companion

About the project →

A public learning-companion demonstration for AI concepts.

Conversational AI · AI literacy · Question answering

AI Workshop Companion

About the project →

A public question-and-answer demo for AI workshop concepts.

Conversational AI · Agentic AI · Question answering

Other publications in this area (1)

Language Models

Building language-model assistants and agents for source-grounded guidance, reflection, and planning.

A knowledge assistant for questions about Utah government records retention.

RAG · Multi-agent systems · Data governance

Related publications (1)

2026 · Conference paper · Project publication

DataGovAI: A Multi-Agent RAG System for Government Data Retention Compliance with Privacy-Preserving Evaluation (opens the publication record in a new tab)

Retrieval-augmented generation · Multi-agent systems · Privacy-preserving evaluation

M. Memari, G. Rudolph, B. Burns, E. Howlett, C. Bramwell, G. Mcewan · IETC 2026 · IEEE

Examines how retrieval-augmented generation and coordinated agents can support government records-retention guidance. The research emphasizes source-grounded responses and privacy-preserving evaluation.

Source-grounded autism information for parents and caregivers.

Conversational AI · Source-grounded QA · Autism education

A conversational companion for everyday reflection and cognitive behavioral skills practice.

Conversational AI · Cognitive behavioral skills · Human–AI interaction

AI Bookkeeping

About the project →

An AI-assisted bookkeeping workspace with a conversational assistant named Saul.

Conversational AI · Financial information systems · Bookkeeping

A finance dashboard that brings balances, spending, and debt information into one workspace.

Data analytics · Decision support · Financial visualization

Other publications in this area (1)

Big Data

Data-intensive analytics and high-performance computing in news-based predictive modeling and computational genomics.

News sentiment & market prediction

About the project →

Explore relationships between news signals and historical stock-market movements.

NLP · Big Data · HPC

Related publications (1)

2017 · Thesis · Project publication

Predicting the Stock Market Using News Sentiment Analysis (opens the publication record in a new tab)

Sentiment analysis · Ensemble learning · Market prediction

M. Memari · M.S. thesis · Southern Illinois University Carbondale

Presented in 2017; the university repository records the award in 2018.

Compares classifiers and ensemble methods using GDELT news/event signals and historical Dow Jones prices from March 2015 to October 2017. This is historical predictive research rather than a live trading system.

Big Data · High-Performance Computing

Other publications in this area (1)

2020 · Journal article

First draft genome assembly of the Argane tree (Argania spinosa) (opens the publication record in a new tab)

Genome assembly · Genomics · Computational biology

S. Khayi et al. · F1000Research, 7, 1310 · Version 2

First published in 2018; revised version published in 2020.

Reports a draft genome assembly for the Argan tree and provides a resource for genomic analysis. The publisher credits Majid with investigation and software contributions.

Big Data · High-Performance Computing