Yousuf Rayhan Emon

About Me

Yousuf Rayhan Emon

AI/ML Researcher | MSc Data Science | Graduate Research Assistant

Research AI/ML XAI

Biography

I am currently pursuing an MSc in Data Science and working as a researcher focused on image-based AI and machine learning. My main research areas include developing lightweight deep learning models for medical imaging and agricultural analysis, using techniques like feature fusion, attention mechanisms, and domain-aware learning. I am also involved in educational data research, especially using structured (CSV/tabular) data to build predictive systems. In all sectors I work in, I apply Explainable AI (XAI) methods to improve transparency and trust in model outcomes. My work has been published in international journals and conferences, and I actively contribute to open-source datasets and real-world AI projects.

Education

MSc in Computer Science & Engineering (Data Science) - Daffodil International University (2024-Cont.) - CGPA: 3.94/4.00
Bachelor's in Computer Science & Engineering - Daffodil International University (2020-2023) - CGPA: 3.87/4.00
Higher Secondary Certificate (Science) - Bangladesh Noubahini School and College, Khulna (2016-2017) - GPA: 4.75/5.00

Experience

Graduate Research Assistant - 4IR Research Cell, Daffodil International University (2024 - Present)
Teaching Assistant - Department of CSE, Daffodil International University (2023 - 2024)
Instructor (Blockchain Development) - Cyber Security Club, DIU (2023)

Research & Technical Interests

Core Areas: Image-based AI, Lightweight Deep Learning, Medical Imaging, Agricultural Analysis
Techniques: Feature Fusion, Attention Mechanisms, Domain-aware Learning, Vision Transformers (ViT)
Specialized: Explainable AI (XAI), Hybrid CNN-ViT Models, Predictive Modeling
Applications: Healthcare AI, Smart Farming, Educational Data Analytics, Blockchain Development

My Projects

DeRefIdm: Decentralized Refugee Management

A blockchain-based system for refugee identification and management using smart contracts. Ensures secure, transparent, and decentralized identity verification.

Details GitHub

AI-Powered Student Dropout Prediction

Machine learning model with explainable AI insights for early detection of student dropout risks. Published in high-impact journal (Citescore 17).

Details GitHub

Smart Agricultural Disease Detection

Vision Transformer models for accurate leaf disease detection in various crops including grape, java plum, and sweet orange with custom datasets.

Details GitHub

Smart Contract Applications

Various blockchain projects including vaccine distribution, insurance management, and secure marriage records using smart contracts.

Details GitHub

Road Damage Detection Using YOLOv10

This project utilizes YOLOv10s to detect and localize potholes, cracks, and other road anomalies with high speed and accuracy, making it suitable for smart city infrastructure monitoring.

Details GitHub

DigitalSignatureWithQRFrontEnd

A React + TypeScript frontend built with Vite, designed for generating and displaying digital signatures embedded within QR codes. This project focuses on a lightweight, modular interface for secure document validation and identity verification workflows. Developed collaboratively with Anupam Abir Kolin, it includes ESLint integration and supports scalable, production-ready code architecture.

Details GitHub

SecureMedical_DIU

A blockchain-based prototype developed using Hardhat to explore secure medical data handling and smart contract deployment. Built in collaboration with Anupam Abir Kolin, the project demonstrates the integration of Ethereum smart contracts for secure, decentralized medical record management. Includes sample contracts, deployment scripts, and basic tests to simulate real-world healthcare use cases.

Details GitHub

MarriageApp-0.02

A decentralized application (dApp) prototype developed using Hardhat to explore blockchain-based marriage registration and validation systems. Created in collaboration with Anupam Abir Kolin, the project features smart contract deployment and testing to simulate secure, tamper-proof documentation of marriage records on the Ethereum blockchain.

Details GitHub

Digital-Signature-Backend

A backend service built to handle digital signature generation and verification processes, designed to complement the frontend QR-based signature system. Developed in collaboration with Anupam Abir Kolin, this project leverages modern cryptographic techniques and API endpoints to ensure secure document authentication in a digital ecosystem.

Details GitHub

Demo_AI_Assistant_DIU

An AI-powered assistant prototype developed to showcase the integration of natural language processing in educational environments. Built for Daffodil International University (DIU), the project demonstrates how conversational AI can assist students and faculty with queries, academic support, and task automation. Features basic interaction capabilities and serves as a foundation for more advanced virtual assistant development.

Details GitHub

Similarity Text Checker with Plotly

A web-based application that analyzes and visualizes text similarity using natural language processing techniques. This tool allows users to compare textual inputs and view similarity scores through interactive Plotly visualizations. Designed for educational and research purposes, it helps identify textual overlap, paraphrasing, or potential plagiarism with clear graphical feedback.

Details GitHub

64_Complex_CNN_Batch_32_Simple_ViT

An experimental deep learning project comparing the performance of a complex CNN architecture with a lightweight Vision Transformer (ViT) on image classification tasks. The models are trained with a batch size of 32, focusing on evaluating efficiency, accuracy, and training behavior. This project serves as a comparative study to explore the trade-offs between traditional convolutional networks and emerging transformer-based vision models.

Details GitHub

SecureMedical_DIU

A deep learning project leveraging a Vision Transformer (ViT) for the classification of leukemia from microscopic blood smear images. This model aims to support early and accurate leukemia diagnosis by capturing complex visual patterns in medical imagery. The project showcases the effectiveness of transformer-based architectures in medical image analysis, with potential applications in automated diagnostics and clinical decision support.

Details GitHub

Deep_Hybrid_CNN_ViT

A hybrid deep learning architecture combining Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) to enhance image classification performance. This project explores the synergistic strengths of CNNs for local feature extraction and ViTs for global context modeling, aiming to deliver robust and accurate visual recognition across complex datasets.

Details GitHub

Research & Publications

Conference Papers

A Fuzzy-Based Vision Transformer Model for Tea Leaf Disease Detection

Authors:Faruk Ahmed, Yousuf Rayhan Emon, Md Taimur Ahad, Mehedi Hasan Munna, Sajib Bin Mamun

Conference: International Conference on Trends in Computational and Cognitive Engineering, 2024

Springer

Scratch Vision Transformer Model for Diagnosis Grape Leaf Disease

Authors:Sajib Bin Mamun, Md Taimur Ahad, Md Monzur Morshed, Nafiull Hossain, Yousuf Rayhan Emon

Conference: International Conference on Trends in Computational and Cognitive Engineering, 2024

Springer

Journal Publications

A Novel AI-Driven Model for Student Dropout Risk Analysis with Explainable AI Insights

Authors:Sumaya Mustofa, Yousuf Rayhan Emon, Sajib Bin Mamun, Shabnur Anonna Akhy, Md Taimur Ahad

Journal: Computers and Education: Artificial Intelligence, 2024 (Citescore 17)

DOI

A Customised Vision Transformer for Accurate Detection and Classification of Java Plum Leaf Disease

Authors: Auvick Chandra Bhowmik, Md Taimur Ahad, Yousuf Rayhan Emon, Faruk Ahmed, Bo Song, Yan Li

Journal: Smart Agricultural Technology, Vol. 8, 2024

DOI

Multi-format Open-source Sweet Orange Leaf Dataset for Disease Detection, Classification, and Analysis

Authors:Yousuf Rayhan Emon, Md Taimur Ahad, Golam Rabbany

Journal: Data in Brief, Vol. 55, 2024

DOI

Data of history: An open-source and multiformat wall image dataset of Panam city, a historical place

Authors:Md Taimur Ahad, Yousuf Rayhan Emon, Sumaya Mustofa

Journal: Data in Brief, Vol. 56, 2024

DOI

BDPapayaLeaf: A dataset of papaya leaf for disease detection, classification, and analysis

Authors:Sumaya Mustofa, Md Taimur Ahad, Yousuf Rayhan Emon, Arpita Sarker

Journal: Data in Brief, Vol. 57, 2024

DOI

Research Under Review

DeRefIdm: A Decentralized Refugee Identification and Management Using Smart Contracts

Status: Under Review at IET Information Security

Description: Blockchain-based solution for secure refugee identity management system.

Multi-Context CNN and ViT Ensemble for Leaf Disease Detection

Status: Under Review at Smart Agricultural Technology

Description: Analysis of image size and batch size effects on ensemble networks.

Open-Source Datasets

BDPapayaLeaf Dataset

Description: Comprehensive papaya leaf dataset for disease detection, classification, and analysis.

Dataset GitHub

Panam City Historical Wall Image Dataset

Description: Multi-format dataset of historical wall images from Panam City for heritage preservation.

Dataset GitHub

Book Chapters

Blockchain Research Framework: A Comprehensive Investigation (Chapter 2)

Book: Practical and Research Applications of Blockchain in Digital Security

Publisher: Muktodhara Prokashoni – DIU Press, 2024

Applications of Blockchain in Refugee Identification and Management (Chapter 10)

Book: Practical and Research Applications of Blockchain in Digital Security

Publisher: Muktodhara Prokashoni – DIU Press, 2024

AI/ML Researcher | Teacher | Python Programmer