StrokeLens
Secure & Explainable Healthcare AI
A secure and explainable AI graduation project focused on data-driven clinical decision support using structured healthcare data.
Tetris AI Lab
End-to-End AI Gameplay Agent
AI agent trained from expert gameplay and strengthened with recovery training, Top-5 one-piece lookahead, and a limited safety filter.
Conceptual project coverPrivacy-Safe Multimodal Authentication
Face, Voice & Speech Verification
A privacy-aware authentication workflow integrating face verification, voice verification, speech recognition, multimodal fusion, and FastAPI backend logic.
- Face verification
- Voice verification
- Multimodal fusion
Dr. NAO
AI-Assisted Dermatology × Robotics
An educational AI and robotics demo integrating a NAO humanoid robot, a web interface, and a pre-trained skin lesion classifier into an interactive workflow.
Medical Image Segmentation
Dental X-Ray Computer Vision
Two applied dental-imaging workflows covering tooth instance segmentation with YOLOv8 and dental-caries semantic segmentation with a custom PyTorch U-Net.
ML Decision Surfaces Lab
Interactive Machine Learning Exploration
An interactive Gradio environment for comparing classification and regression models through decision boundaries, ROC-AUC, confusion matrices, learning curves, and controlled noise experiments.
Imaging Filters Lab
Interactive Digital Image Processing
An interactive Digital Image Processing lab built with OpenCV and Gradio for exploring spatial filters, FFT, wavelets, morphology, noise, restoration, and binary operations.
Intel Image Classification
Transfer Learning · EfficientNetB0 · Fine-Tuning
A natural-scene classification project comparing a custom CNN with EfficientNetB0 transfer learning and fine-tuning across six image classes.
Handwriting OCR
HOG Features · KNN Classification · EMNIST Letters
A classical OCR pipeline for handwritten English letter recognition using HOG feature extraction and a distance-weighted KNN classifier, extended with character segmentation and custom handwritten word reconstruction.
MNIST CNN Experiments
Controlled Deep Learning Comparison
A controlled PyTorch comparison of CNN architectures for handwritten digit classification, focused on the effect of Batch Normalization on accuracy, loss behavior, runtime, and prediction quality.
Blood Donation System
Donor Matching · Blood Requests · Admin Management
A PHP + MySQL web application for donor registration, blood-type compatibility search, blood-request management, and authenticated admin workflows.
ADDITIONAL KAGGLE EXPERIMENTS
Focused experiments beyond the main case studies.
Smaller, targeted notebooks kept intentionally separate from the full project showcase so the main portfolio stays focused while still documenting broader ML practice.
A structured classification experiment covering EDA, preprocessing, categorical encoding, scaling, train-only SMOTE, and comparison of multiple machine-learning models.
- SMOTE applied only after train/test split
- Multi-model comparison
- Precision / Recall / F1 evaluation
A focused NLP baseline for spam detection using cleaned text features, stratified evaluation, and Multinomial Naive Bayes with class-aware performance reporting.
- Text preprocessing
- Stratified train/test split
- Confusion matrix + classification report