Innovation Starts Here
Empowering the next generation of innovators through collaborative capstone projects. Join a community where creativity meets technology.
Featured Projects
Explore innovative solutions created by our talented students across various domains and technologies.
Automated Osteoporosis Detection Via Ensemble based Hybrid Deep Learning Approach
Automated osteoporosis detection using ensemble-based hybrid deep learning has emerged as a high-performance, non-invasive method for early diagnosis, frequently achieving accuracies exceeding by leveraging multiple AI architectures simultaneously.
Diagnosis of Alzheimer’s Disease from MRI using Convolutional Neural Network (CNN) Model.
Convolutional Neural Networks (CNNs) automate Alzheimer’s disease diagnosis from MRI, achieving high accuracy by detecting structural brain changes, particularly in the hippocampus, which signify early-stage dementia
Automated Road Damage Detection and Classification in Bangladesh Using Transfer Learning
Road damage classification in Bangladesh using deep learning, specifically CNN-based transfer learning models, achieves high accuracy in identifying potholes, cracks, and surface degradation from street-level images.
Osteoporosis Knee Detection Using X-Ray Images via a Hybrid DenseNet169–Vision Transformer Feature Fusion Ensemble
A hybrid deep-learning model, often referred to as BONE-Net, utilizing a combination of DenseNet169 and Vision Transformer (ViT) with an Attention Model (AM), has demonstrated high efficiency in detecting osteoporosis from knee X-ray images.
Multi-Task Neural Networks for Glioma Classification and Survival Prediction Using High-Dimensional Gene Expression Data
Multi-task neural networks are increasingly used to address the high dimensionality and complexity of gene expression data for glioma analysis, enabling simultaneous classification of subtypes and prediction of patient survival.
Dual-CNN Region-Level Feature Fusion for Image Captioning Using Transformer Decode
Dual-CNN region-level feature fusion for image captioning using a Transformer decoder is a sophisticated computer vision-natural language processing (CV-NLP) approach designed to improve descriptive accuracy by combining distinct visual representations.
Ready to Start Your Innovation Journey?
Join hundreds of students who have transformed their ideas into reality. Your capstone project awaits.