My Research Works
A showcase of my research contributions
AllAgricultural AIBangla Language ProcessingBiomedical Signal AnalysisComputer VisionDeep LearningExplainable Artificial Intelligence (XAI)Fuzzy Ensemble LearningHybrid DatasetsLow-Resource Language ProcessingMedical Data AnalysisNamed Entity Recognition (NER)Natural Language Processing (NLP)Regional Dialect AnalysisSignal ProcessingTransfer LearningVision Transformer (ViT)
VR-FUSENET: A FUSION OF HETEROGENEOUS FUNDUS DATA AND EXPLAINABLE DEEP NETWORK FOR DIABETIC RETINOPATHY CLASSIFICATION
May 30, 2025
This paper proposes VR-FuseNet, a hybrid deep learning model combining VGG19 and ResNet50V2 with explainable AI techniques for accurate and interpretable diabetic retinopathy detection.
Computer VisionDeep LearningTransfer LearningExplainable Artificial Intelligence (XAI)+2
- Hybrid dataset integration from five public sources
- Effective handling of dataset imbalance using SMOTE
Butterfly Optimization and Deep Learning to Classify Heart Sound Signal
May 9, 2025
This paper proposes a non-invasive PCG signal-based classification method using deep learning models optimized with the Butterfly Optimization Algorithm, achieving 99.07% accuracy in cardiovascular disease detection.
Signal ProcessingDeep LearningBiomedical Signal AnalysisMedical Data Analysis
- Non-invasive PCG signal-based classification method
- Feature extraction using MFCC and WST
Transfer Learning and Explainable AI for Brain Tumor Classification: A Study Using MRI Data from Bangladesh
April 11, 2025
This study presents an automated brain tumor classification system using MRI data from Bangladesh, leveraging transfer learning and explainable AI to achieve 99.17% accuracy in detecting glioma, meningioma, and other brain cancers.
Medical Data AnalysisDeep LearningTransfer LearningExplainable Artificial Intelligence (XAI)
- Developed an automated MRI-based brain tumor classification system
- Utilized VGG16, VGG19, and ResNet50 models for tumor classification
ANCHOLIK-NER: A Benchmark Dataset for Bangla Regional Named Entity Recognition
March 14, 2025
ANCHOLIK-NER introduces a linguistically diverse dataset capturing Bangla regional dialects across various regions for advancing Named Entity Recognition (NER) research.
Natural Language Processing (NLP)Named Entity Recognition (NER)Bangla Language ProcessingLow-Resource Language Processing+1
- Creation of the first large-scale Bangla regional dialect NER dataset
- 17,405 manually annotated sentences across five regions
An Exploratory Approach Towards Investigating and Explaining Vision Transformer and Transfer Learning for Brain Disease Detection
December 22, 2024
This study presents a comparative analysis of Vision Transformers (ViT) and Transfer Learning (TL) models for classifying brain diseases from MRI scans in Bangladesh, enhanced by Explainable AI techniques.
Vision Transformer (ViT)Transfer LearningMedical Data AnalysisDeep Learning+2
- Conducted comparative analysis between ViT and TL models for MRI-based brain disease detection
- Used Bangladeshi MRI dataset to train and evaluate model performance
Comprehensive Lung Disease Detection Using Deep Learning Models and Hybrid Chest X-ray Data with Explainable AI
December 22, 2024
This study evaluates deep learning and transfer learning models for lung disease detection using hybrid chest X-ray datasets and employs Explainable AI to enhance interpretability and clinical trust.
Medical Data AnalysisHybrid DatasetsDeep LearningTransfer Learning
- Merged four chest X-ray datasets including two from Bangladesh
- Created a hybrid dataset for more generalizable training
Fuzzy Rank-Based Ensemble Learning for Eye Disease Classification Using Retinal Images: A Bangladeshi-Specific Dataset with Explainable AI Integration
December 22, 2024
This paper presents a fuzzy rank-based ensemble approach for retinal disease classification using Bangladeshi datasets, integrating Explainable AI to enhance model interpretability and medical trust.
Fuzzy Ensemble LearningMedical Data AnalysisHybrid DatasetsDeep Learning+1
- Collected region-specific retinal images from Bangladeshi health facilities
- Evaluated multiple DL and TL models for eye disease classification
An Approach Towards Identifying Bangladeshi Leaf Diseases through Transfer Learning and XAI
December 20, 2024
This study proposes an accessible deep learning solution to classify 21 distinct leaf diseases across six plant species in Bangladesh using CNN, transfer learning, and explainable AI techniques.
Agricultural AIDeep LearningTransfer LearningExplainable Artificial Intelligence (XAI)
- Classified 21 leaf diseases across six plant types
- Used fine-tuned deep learning models like VGG19, MobileNetV2, and ResNet50V2
A Comprehensive Analysis of COVID-19 Detection Using Bangladeshi Data and Explainable AI
October 26, 2024
This paper proposes an explainable deep learning approach for detecting COVID-19 from chest X-rays using a Bangladeshi dataset, emphasizing transparency, class balancing, and clinical interpretability.
Medical Data AnalysisDeep LearningTransfer LearningExplainable Artificial Intelligence (XAI)
- Utilized a Bangladeshi dataset of 4,350 CXR images
- Classified images into four categories: Normal, Lung-Opacity, COVID-19, and Viral-Pneumonia
Advanced CNN and Explainable AI Based Architecture for Interpretable Brain MRI Analysis
October 18, 2024
This paper proposes an interpretable CNN-based architecture for brain MRI analysis, integrating Explainable AI (XAI) techniques like LIME to improve model transparency and aid clinical decision-making.
Medical Data AnalysisDeep LearningTransfer LearningExplainable Artificial Intelligence (XAI)
- Used a dataset of 5,285 brain MRI images
- Achieved classification accuracy of 86%
Improving Bangla Regional Dialect Detection Using BERT, LLMs, and XAI
September 25, 2024
This paper presents a novel approach for identifying and categorizing Bangla regional dialects using transformer-based models like Bangla BERT and GPT-3.5, combined with explainable AI techniques.
Natural Language Processing (NLP)Bangla Language ProcessingLow-Resource Language ProcessingRegional Dialect Analysis
- Collected a diverse dataset of Bangla regional speech samples
- Applied Bangla BERT for dialect classification with 88.74% accuracy