Engineering/Science Fairs

I help volunteer, mentor, and judge students for high school science fairs, science olympiad, and other competitions. Recently, I was a judge at Synopsis Science Fair in 2015. Previously, I mentored several successful teams of science fair students at these contests from 2012-2014. Students I have worked with have placed at the international level at Google Science Fair, the state level at Synopsis, and featured in several local periodicals. So proud of them! Check out some of their projects below!

HAWK: Anti-Hate and Harassment Classifier (2016)

Handwashing Hound (2015)

Parking Pigeon (2014)

Read more about it in the Mercury News:

Wi-Fi Watchdog (2012, 2013)


In college I competed in several hackathons. I also helped put on a few on as part of the USC Association for Computing Machinery (which I was a founding officer my freshman year). I was the lead organizer of the first USC vs. UCLA SS12 programming contest. These programming contests are now annual events (and the ACM team has won Best Viterbi Engineering Student Organization twice in eight years!).

These days I help mentor/judge students in Robotics, Machine Learning, Android, and general programming for high school science fairs and entrepreneurial contests. I am always interested in helping people learn to code and build ever more intelligent machines. I am extremely excited that computer science is becoming more of a science and increasingly contributing to science too!


I was a Speech and Debate captain in high school for Lincoln-Douglas debate. I still actively judge/coach high-school debaters from time to time. I am always looking for volunteer judging opportunities in the activity.

In addition, I used to work on MVLDDB, a database of high school evidence cards for my Speech and Debate team. I later studied automated summarization algorithms in natural language processing, which at some point needs to be brought back to debate.

Finally, check out blue prints for an ambitious robotic debater project!

The debate roombas in action:


I’m a big fan of many different types of music. I learned piano for 5 years, played clarinet in my high school marching, symphonic, and jazz bands, and have been learning guitar since freshman year in college.

Here are some recordings from Summer 2014 from my SoundCloud account mostly on guitar and guitar-driven midi-instruments

I took part with my roommates Henry Yuen and Noel Overkamp in the Viterbi Talent Show in undergrad back in 2009 where we performed “Song of Storms” from Zelda 64. Watch my last clarinet performance EVER here:

After the performance, I was regrettably banned from ever touching the instrument again since thunder and lightning struck Los Angeles (and it never rains in Los Angeles). For fear of the mystic power of the clarinet and the regrettable fact that it drove too many of our neighbors to complete insanity, I have quietly retired from my clarinet days. I have henceforth transitioned to my sound cloud abode with some original audio hacks on electric guitar, midi instruments, etc.

My interest in learning music has also translated into several useful guitar and music apps. Check out some of my guitar apps and melody recognition open-source software!

Guitar Stylometry from Motion:

Intelligent Accessible Guitar Tab Reader:

Real-Time Melody Recognition System:


I enjoy playing virtually any sport with a racket: tennis, badminton, table tennis and am always up for a game or two.

I run the CMU RI Table Tennis League and also organize table tennis tournaments from time to time. Let me know if you want to play me! Or my robot. Be warned, the robot is quite good:

Oh, and USC Football. Fight On! =]


Even more hack home projects:

Home Aquarium Monitor

I have a bunch of web cams around so decided to try something “useful” with them, while shamelessly playing around with some visual processing techniques in the process. I’ve been trying to build a monitoring system for my home aquarium. It’s doing pretty decent job for a first take.

This is a four web cam set up around a 1.5 gallon aquarium. Two cameras are on the top of the aquarium and two are mounted on the side. The four rows on the screen show the output of the different cameras as well as some visually processed results. The screen shots from left to right are:

(1) The raw output of the webcam.
(2) Basic motion detection in the scene. Pretty much does a diff between the current image and last two images.
(3) Edge Detection – looks for differences in color in neighboring pixels.

Alas, some of the cameras are doing a lot better than others at isolating the fish. Watch camera 2 and the motion detection screen for it and you can see its doing a pretty decent job of isolating the fish in the scene.

I hope to eventually not only improve the the fish motion tracking (perhaps using more temporal and the edge data) but get the cameras to collaborate to figure out where each of the fish are likely to be in the tank for each time step.

Guitar Melody Extraction and Classifier System

I tried to build system to allow a human guitar player to hook up their guitar to their computer, and recognize melodic segments in real-time. The system used chroma features as well as a bunch of other audio signal processing techniques. I built the system as a hobby project during Summer 2010. See links:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s