Research
Research
Data Mining
Computer Vision
Health Informatics
Natural Language Processing
Machine Learning
Human-computer interaction (HCI)
Ongoing Projects
AI-Based Clinical Recommendation System
Developed a machine learning-based model to reduce antibiotic resistance and irrational medication based on a patient's symptoms and demographic data to predict the most appropriate course of action, either over-the-counter (OTC) medications or a doctor’s consultation. Designed a diabetes prediction model tailored to the Bangladeshi demographic using machine learning and data mining techniques.
Supervisor: Dr. Khondaker Abdullah-Al-Mamun, Professor, Dept. of CSE & Director, IRIIC, UIU, Bangladesh
Big5 Personality Trait-Based Contextual Gamification Platform
Selection Identified contextual gamification platforms for accurate personality trait extraction. Explored open-vocabulary and dictionary-based approaches using LIWC and word embedding techniques. Utilized BERT and NLP features for personality prediction.
Supervisor: Dr. Md Saddam Hossain Mukta, Post-doctoral Researcher, LUT University, Finland
Undergraduate Thesis
Undergraduate Thesis
The title of my thesis is “ A Bidirectional Siamese Recurrent Neural Network for Accurate Gait Recognition Using Body Landmarks.” Our method leverages advanced techniques for gait recognition, including sequential gait landmarks obtained through the Mediapipe pose estimation model, Procrustes analysis for alignment, and a Siamese biGRU-dualStack Neural Network architecture for capturing temporal dependencies.
Developed a new approach for gait recognition from real-time video, accurately identifying individuals based on unique walking patterns.
Leveraged Mediapipe pose estimation model for obtaining comprehensive gait cycle coverage and enhancing gait pattern representation.
Achieved high recognition accuracies on large-scale cross-view datasets and outperformed state-of-the-art methods in biometric identification through gait analysis.
Contributed to the field of gait recognition with potential applications in security, surveillance, and healthcare domains.
Supervisors:
Professor, Dept. of CSE
BRAC University, Bangladesh
Director, Brain Innovation
Research Experience
Malaria Cell Classification using Deep Learning Models Jun, 2022 – Jan, 2023
Developed deep learning models for identifying and forecasting malaria-infected blood cells. Achieved performance
comparable to existing methods for thin blood smear RBC slide images, showing potential for malaria detection and similar
diseases.
Supervisor: Dr. Swakkhar Shatabda, Professor, Dept. of CSE, BRAC University, Bangladesh
Comparative Analysis of Protein Secondary Structure Prediction Jul, 2022 - Nov, 2022
Conducted comparative analysis of protein secondary structure prediction methods using a graph neural network (GNN).
Achieved accurate predictions using SVM, Adaboost, KNN, and Decision Tree models, demonstrating the superiority of the
traditional approach.
Supervisor: Subangkar Karmaker Shanto, Graduate Student (PhD in Computer Science ), Purdue University, USA
AI-Based Squirrel Detection System for Crop Protection Sep, 2022 - Mar, 2023
Developed an AI-based system using YOLOv5 with a SENet attention layer to detect squirrel variants in real-time, protecting
crops. Collected and labeled video data from fields and social media. Implemented a cloud-based AI system with cameras for
real-time monitoring and validation across multiple locations.
Supervisor: Dr. Dewan Md. Farid, Professor, Dept. of CSE, UIU, Bangladesh
See more in Google Scholar & ResearchGate