About Me

I’m a machine learning researcher and high school student in San Diego, California. Although I’m still in secondary school, I have a passion for academic research and have published three papers in various conferences of machine learning, computer vision, and robotics. I have some experience in industry too, acting as a software developer for a startup company. In my free time I enjoy reading Stoic philosophy, stargazing, singing Indian classical music, and going on walks with my dog Peanut.

Download Resume


MultiNet: Mulit Modal Deep Learning for Autonomous Driving

Autonomous driving requires operation in different behavioral modes ranging from lane following and intersection crossing to turning and stopping. I created a technique for learning multiple distinct behavioral modes in a single deep neural network through the use of multi-modal multi-task learning. I tested this technique using labeled data from over one hundred hours of driving our fleet of 1/10th scale model cars.

PhonoNet: Raga Predicton and Visualization

Indian classical music is a fading art form which focuses predominantly on improvisations on specific melodic modes, or ragas. The improvisational nature of the Indian ragas make them very difficult to consistently record in a distributable written format. Because ragas are passed down solely through an ancient oral tradition, much of the rich knowledge of Indian ragas is dying.

To address this issue, I created the PhonoNet deep learning system for raga detection, a tool which documents and augments the learning of Indian classical music and ragas by providing raga prediction and audio visualization.

View on GitHub

SimpleCV for Computer Vision Workshop

As the president of my high school robotics team, I lead a number of outreach efforts. For one of our team’s STEM workshops, I created a simple wrapper on top of OpenCV called SimpleCV to teach middle school students the basics of image manipulation and computer vision. At the end of the activity, all students could draw a mustache on top of an image of a face and some were able to draw mustaches on themselves with a live feed camera!

View on GitHub

Haar Cascade Training Library on Windows

A comprehensive library for training Haar Cascades on Windows machines (primary operating system of my robotics team). All necessary files are compiled and stored within the repository with detailed usage instructions. Don’t get me wrong, Linux RULES! This is just so my robotics computer vision team members could get started.

View on GitHub