AKNotes

A secure Android application for note-taking, incorporating user authentication via Firebase. It uses CRUD operations for notes, stored in Firestore to allow seamless cloud storage maintaining data uniformity across devices.

AndroidJavaFirebaseFirestoreUser AuthenticationMVC
Bengal Iron Calculator

This Android App connects with Firebase to dynamically fetch product data, weight and price per unit weight to display it in a list view format. Users can search for products and add products to cart by weight and price and it calculates the total price and weight for the entire cart. This app uses advanced filtering and search functionalities to enhance user experience, and leverages the MVVM design pattern to ensure efficient data management and separation of concerns. The solution significantly streamlined the client’s sales operations.

AndroidKotlinFirebaseFirestoreMVVM
Semantic Segmentation of Point Clouds for 3D Object Feature Recognition

Developed a machine learning model for a CPSC 440 Advanced ML and AI course to predict surface features of 3D objects from Point Clouds. We tried different clustering algorithms on surface normals of point clouds to produce features for the linear classifier to predict surface features of 3D objects. Designed the architecture and implemented the training and prediction of the Linear Classifier model in PyTorch. Explored literary research in subject area and assessed each method along with the results to compare with our use case.

Machine LearningPyTorchDeep LearningFeature Extraction
Socket Chat Web App

Utilized NodeJS and ExpressJS to establish a robust backend and socket.io to faciliate real-time bi-directional event-based communication. This web-app supports concurrent connections, handles message broadcasts, and includes user authentication mechanisms to maintain session integrity.

Node JSJavaScriptHTMLCSSSocket.io