CV/Resume

Graduate Research Assistant

Michigan State University

Working in iPRoBe Lab under the supervision of Dr. Arun Ross.

• Explainability in Face Recognition
• Visualization Applications
• Forensic Biometrics

Graduate Teaching Assistant

Michigan State University

Conducted lab sessions of CSE-102 Algorithmic Thinking and Programming in Fall-2021 and Spring-2022 semester.

• Algorithm Development
• Python Programming

Lecturer

Islamic University of Technology, Gazipur, Bangladesh

Conducted regular undergraduate level courses and held lab sessions along with other official duties.

• Face Recognition
• Biometric Template Protection
• Deep Learning

Lecturer

Daffodil International University, Dhaka, Bangladesh

Conducted regular classes and lab sessions.

• C++ programming
• Database Management

Intern

XeonBD Limited

Acquired skills in working with web technologies, including hosting and maintaining websites through cPanel

• Web Hosting
• cPanel

General Secretary

Bangladesh Student Association, Michigan State University

A community bound by rich Bangladeshi culture.

• Community leadership & Engagement
• Event Planning
• Public Speaking

Doctor of Philosophy

Michigan State University

Computer Science and Engineering

Master of Science

Michigan State University

Computer Science and Engineering

Bachelor of Science

Islamic University of Technology

Computer Science and Engineering

Higher Secondary Certificate

Pabna Cadet College

Science

Secondary School Certificate

Pabna Cadet College

Science

Best Poster in MSU Engineering Graduate Research Symposium-2024

College of Engineering, Michigan State University

For the poster on 'Comparative Medical Radiography using Deep Neural Networks'.

Fully Funded OIC Scholarship for Bachelor's Degree

Organization of Islamic Cooperation (OIC)

Awarded based on the admission test result which covers all the cost for 4-year Bachelor of Science degree.

A Parametric Approach to Adversarial Augmentation for Cross-Domain Iris Presentation Attack Detection

WACV-2025

This study introduces ADV-GEN, a convolutional autoencoder that generates adversarial samples by applying geometric and photometric transformations to both genuine and PA irides. Using these samples, the proposed approach enhances the cross-domain performance of presentation attack detection (PAD) classifiers, as demonstrated on the LivDet-Iris 2017 database.

Investigating Weight-Perturbed Deep Neural Networks With Application in Iris Presentation Attack Detection

WAACV-2024

This study analyzes the sensitivity of three DNN architectures (VGG, ResNet, DenseNet) to parameter perturbations, using techniques like Gaussian noise, weight zeroing, and weight scaling on the LivDet-Iris-2017 and LivDet-Iris-2020 datasets. The proposed approach combines perturbed models at score and parameter levels, achieving average performance improvements of 43.58% and 9.25% on the respective datasets.

Less is More: Lighter and Faster Deep Neural Architecture for Tomato Leaf Disease Classification

IEEE Access

This study presents a lightweight, transfer learning-based model using MobileNetV2 for disease detection on tomato leaves, enhanced by illumination correction and runtime data augmentation to prevent leakage and address class imbalance. Tested on the PlantVillage dataset, the model achieves 99.30% accuracy with only 9.60MB size and 4.87M operations, making it highly suitable for low-end devices.

Projects Page

This will direct to the projects page of my personal website.