Hi, I'm Sri! I work in game design as the founder of Zenovia Interactive. In a past life I worked at Bloomberg LP as a finance engineer, and before that I researched artificial intelligence at the University of Maryland.
Steel Assault is my current WIP videogame
project. It's a fast paced 2D platformer revolving around
close combat and a variety of unique weapons. I'm programming,
designing, and composing the music for this game. The art is by
For MHacks Winter 2014, our group combined the
Oculus Rift virtual reality headset and Leap Motion motion
sensor to create a basic 3D first person shooter. The Oculus
Rift is used to look around and "experience" a virtual world,
and the Leap Motion is used to navigate it through hand
and runs entirely in-browser, utilizing three.js,
vr.js, and jQuery.
For the final project of my Computer Science Honors Seminar
class, my group created an interface using the Microsoft
Kinect where the user could create, drag, resize and play
virtual drums in physical space using gestures. The pitch
of each drum varied with its size. I worked mainly on the
sound portion of this project, creating event handlers for
when a drum was hit which repitched and played a drum sound
on the fly.
Age related macular degeneration (AMD) is the leading cause of blindness
for people over the age of 50 within much of the Western world. During my
internship at the Johns Hopkins University Applied Physics Lab, I designed
and implemented an algorithm to automatically detect and classify the severity
of AMD from retinal images. The algorithm achieved a sensitivity of 98.6% and
a specificity of 96.3%, when compared with expert fundus gradings for 468
patients. As of September 2015, this is the highest published accuracy rate for
automated diagnosis of AMD.
A simple robotic system capable of automated navigation and rudimentary object recognition,
based off the Videre Erratic robot. After training, the robot was able to successfully scan
its surroundings for a given object in its knowledge set, then triangulate the position of
that object in real space and navigate to it (avoiding any obstacles in its path). This
project was shown at the Baltimore Science Fair where it won several awards, including the
Intel Excellence in Computer Science Award and the US Army Public Health Command Award of Merit.
A computer program which uses a Kohonen neural network for handwriting and font
recognition (OCR), focusing on individual letters. During experimental testing, the program
was successfully able to recognize 90% of the font images and 80% of the handwritten images
fed to it. It was submitted to the Baltimore Science Fair where it won several awards,
including the NSA Outstanding Use of Mathematics Award and the ACM Excellence in Computer