Projects

leader-follower

Collaborated with a 5-member team on computer vision for leader-follower tracking with automatic obstacle detection and coordination for seamless platoon movement.

Autonomous Aerial Filming

As part of a team of three, we developed an architecture capable of emulating realistic aerial filming through YouTube video dataset demonstrations and autonomous drone navigation. My role focused on researching and implementing a deep learning framework for camera pose prediction.

Optimal Robotic Control

In addressing state dynamics and input uncertainty in robot manipulation, we implement an LQR controller from the Optimal Control Approach to solve the robust control challenge.

Implementation of UnDeepVO

The Unsupervised Deep Learning Visual odometry research paper was Implemented in Tensorflow to check the capability of Model in real world dataset, using just monocular camera.

New Eye

This project was created back when I was doing undergrad, to help blind people move around the world more safely.

Search and Rescue Drone

This Project was developed to help the rescue teams to search the areas natual calamities and detect humans using AI and drone.

Irona

Elevate your home to a child-friendly, pristine haven with 'Irona,' our robot maid driven by Computer Vision and ROS (the Turtlebot) with movebase planning. Simply place it in the room, initiate the cleaning, and observe Irona scanning the floor, recognizing toys using Aruco markers, and orchestrating an efficient tidying process.

Robo-Serve

Incorporating knowledge from the Master's in Robotics courses (ENPM662 and ENPM661), this custom project caters to the hospitality industry's requirements. It employs the RRT* planning algorithm on uniquely designed wheeled robots, turning them into autonomous servers.

Autonomous Robotic Arm

The Project was based on Image processing and 3 DOF cartesian manipulator, to make fully autonomous robotic arm.

LQG Controller for Moving Crane

This project was develop as a part of final project for Control System (ENPM667) course in Master's in Robotics, and the task was to add LQR and LQG Controller for double pendulum car problem.

IOT Web Server

After watching Iron Man mutliple times, decided to build own home automation system without using cloud server, hence the IOT server was created, with all the capability of server plus the ability to communicate to IOT devices in its vicinity.