Developed a web-based second-opinion platform using deep learning to analyze chest X-rays and detect 14 chest conditions, achieving 92% AUROC. Implemented real-time inference with Grad-CAM heatmaps and likelihood-based predictions for enhanced diagnostic accuracy.
Developed CNN, SVM, and KNN models to classify fire and non-fire images from CCTV footage, achieving 97% accuracy with CNN. This classification enables real-time fire detection, enhancing safety by quickly identifying fire incidents in CCTV footage.
Geared towards 5th and 6th graders, this app simplifies concepts of perimeter, area, and angles through captivating lessons and an exciting adventure game that progresses as students answer quizzes.
A game that embodies OOP principles - encapsulation, inheritance and polymorphism. With an AI that plays and blocks winning moves, it offers a user-friendly interface, turn-based gameplay, and diverse modes while showcasing an engaging experience.
Two-player Connect 4 game in Python, featuring object-oriented design with classes for game pieces and boards. Users can create custom instances. Robust exception handling ensures smooth gameplay by preventing crashes from invalid inputs.
A Hospital Management System, built using Python and SQL database, it stores details of doctors, nurses, patients and manages appointments.
