I am a Computer Science PhD student at Arizona State University’s REACT lab where, under the guidance of Professor Stephanie Gil, I am exploring an application of data from different senors (laser, camera) extracted from robot platforms for addressing SLAM. I am also working on using the underlying communication network based on WiFi, as a sensing device with application in robotics. As part of this research I have had hands-on experience on framworks like, ROS and GTSAM, robotics platforms like TURTLEBOT3 and CREATE2 as well as Opti-Track Motive. I am also a freelance consulatant for SENSAGRATE where I advise the development team on system architecture and framework for building applications using data from Lidar and Camera. I was an intern at 10 Imaging Inc during summer 2018 where I developed an application that identifies symbols (language, mathematical) extracted from HD visual input.
You can checkout the details of my other projects below.
Deep Learning, Artificial Intelligence and Robotics:
- I have implemented some deep networks like Autoencoders using Keras/ Tensorflow, a numpy implementation of Multi-class neural networks and analysed reinforcement learning algorithm using openAI gym’s MsPacman environment.
- For my robotics project I implemented an IEEE paper that deals with Formation fight control for a multi - UAV system using fourth order dynamics and Model Predictive Control and developed a system that enables indoor localization of robots using WiFi
Machine Learning
- These include recommender systems for datasets like YELP and IMDB,MovieLense etc and implementation of generic supervised and semi-supervised machine learning algorithms like linear and logistic regression, K Nearest Neighbours, K-means etc.
- Used augmented reality to extract usage behavior of users while playing with an application that teaches Java language. The goal was to enable the application to self-adapt for each user.
Systems and Security
- Projects include linux systems calls, kernel programs and an application of a distributed database optimized for queries targeted towards geolocation.
- Building a complete framwork in AWS elastic cloud that optimizes virtual machine usage based on generated requests and performs object detection on camera images collected from RPi cameras mounted on ground robots.