Projects

Payment Detail Register

August 2024

  • Developed a single-page application to manage payment details (card owner name, card number, CVV, and expiration date) using C#, .NET Core for backend services and Angular for the frontend.
  • Implemented functionalities for adding, updating, and deleting card entries in a SQL database, with validation on all fields (16-digit card number, 3 or 4-digit CVV, 5-character expiration date).
  • Enhanced the user experience by applying form validation, showing invalid fields with a red border, and disabling the submit button for incomplete or incorrect entries.
  • Integrated ngx-toastr for real-time notifications on success, failure, and update actions, ensuring user-friendly feedback.
  • Secured sensitive card data using best practices for input validation and error handling.

MovieShop MVC Application

July 2024

  • Developed a comprehensive MovieShop MVC Application utilizing ASP.NET Core with Onion/Clean Architecture to implement a robust and scalable web application.
  • Implemented CI/CD pipelines using Azure DevOps, ensuring seamless integration and deployment processes. Automated build and release processes, reducing deployment time by 50%.
  • Configured Azure SQL Database for backend storage, leveraging Entity Framework Core for data access and migrations, resulting in efficient data management and operations.
  • Enhanced application performance and scalability by deploying the application on Azure App Service, optimizing resource utilization and ensuring high availability.
  • Integrated user authentication and authorization using ASP.NET Core Identity, providing secure access control and user management.
  • Set up a dynamic genre-based movie browsing feature, fetching and displaying movies by genre, improving user experience and engagement.
  • Utilized Bootstrap for responsive design, creating a user-friendly interface that works seamlessly across various devices and screen sizes.
  • Conducted thorough testing and debugging of the application, ensuring high-quality and bug-free releases through automated testing in the CI/CD pipeline.

Optimizing Energy Efficiency: A Comprehensive Analysis of Household Electricity Consumption

Mar 2024 - May 2024

  • Developed machine learning models to predict household electricity consumption, using time-series analysis to enable data-driven decisions for energy savings. Implemented various models including Polynomial Regression, XGBoost, Random Forest, and CatBoost, with Random Forest achieving the lowest RMSE.
  • Enhanced model insights by integrating findings into a user-friendly interface using the Django framework, facilitating real-time energy consumption predictions to promote efficient energy use and cost savings

Robotics Project

Sept 2023 - Nov 2023

  • Developed and implemented a quadruped navigation and obstacle avoidance system for a robotic dog using the Unitree Go1 robot in the Isaac Gym Simulator.
  • Utilized the physics-based simulator, Isaac Gym, for environment setup, interfacing with the gym environment, and training neural networks through model-free reinforcement learning.
  • Successfully achieved the project’s primary objective of enabling the robot dog to navigate from point A to point B while dynamically avoiding collisions with obstacles, showcasing expertise in legged robot control and reinforcement learning

Age, Gender, Ethnicity Prediction

June 2023

  • Developed an image classification system using deep learning techniques to predict age, gender, and ethnicity from uploaded images.
  • Implemented a ResNet9 architecture for gender and ethnicity prediction, and a ResNet18 architecture for age prediction.
  • Trained the models on a labeled dataset containing images representing various age, gender, and ethnicity groups.
  • Developed a Flask web application to provide an interactive user interface for image upload and prediction.
  • Integrated the trained models into the Flask app, allowing users to upload an image and receive predictions for age, gender, and ethnicity.

Two-Way Real-Time Sign Language Recognition

Feb 2023 – Apr 2023

  • Developed a Two-Way Real-Time Sign Language Recognition system using a Convolutional Neural Network (CNN) for American Sign Language (ASL) and Indian Sign Language (ISL).
  • Used a dataset with 140 images for each sign for both language systems, used skin detection and hand segmentation techniques to isolate the hand region from the background.
  • Implemented CNN for classification of signs and achieved an accuracy of over 90% on the test set.

American Sign Language Recognition using Yolov5

Feb 2023 – Apr 2023

  • Developed a Real-Time American Sign Language Recognition model using Yolov5.
  • Annotated 40 images for each sign using Roboflow to prepare the training and validation dataset.
  • Received a confidence level of above 0.7 on all signs while testing.

Fast Trajectory Replanning

Sep 2022 – Oct 2022

  • Led a team of 3 in developing a maze-solving application using the A* search algorithm that enables an agent to navigate a 101x101 grid maze with obstacles to reach a target cell using the shortest route, with a Manhattan distance heuristic.
  • Implemented and tested backward A* search and adaptive A* search algorithms, comparing their execution times to forward A* search.
  • Created an interactive graphical user interface using the pygame module to visually display the shortest path the agent takes to reach the target.

BidBazaar

Feb 2023 – Apr 2023

  • An auction website like eBay, with features like user account creation, auctions, browsing and advanced search functionality, and admin and customer representative functions.

Battery Management and Data Analytics of Battery and Vehicle Data

Sep 2021 – Apr 2022

  • Led a team of 4 in designing a Data Acquisition System and Battery Management System and integrating it into the brain of the vehicle, (Vehicle Control Unit) to create a robust electronics system for an Electrics Vehicle.
  • Prepared Machine Learning models such as Gradient Boosting, Random Forest, LASSO Regression to predict the SOC (State of Charge) of the battery of the vehicle. Gradient Boosting achieved the best R2 score of 0.977.

Electrokeet

Sep 2019 - May 2021

  • Our Team’s entry to IICDC 2019-2020 competition - Electrokeet is a drone that uses thermal imaging and AI to detect pests at their nascent phases of life.
  • Integrated AMG8833 with Arduino UNO to design a thermal camera.
  • Used a Texas Instruments board, MSP430F5438 and integrated it to a soil sensor.
  • Designed the front end of the market place of the website using HTML, CSS, JS.

Unmanned Autonomous Vehicle

Sep 2020 - Apr 2021

  • Engineered an unmanned autonomous vehicle to improve the disaster management system of our country, by delivering food and medicines from time to time to the numerous quarantine centers.
  • Performed a running simulation of the autonomous robot on Gazebo using ROS. Executed JS framework to design a GUI and assembled a prototype using Raspberry Pi 3B, Rpi Cam, Arduino UNO, HCSR-05 ultrasonic sensor and GPS module.