[Super Student]Jay Rambhia: The computer vision whiz-kid.

By Ganesh Nagarsekar|18th Jun 2013
Clap Icon0 claps
  • +0
    Clap Icon
Share on
close
Clap Icon0 claps
  • +0
    Clap Icon
Share on
close
Share on
close

Like thousands of teenagers w

grav2

ho find themselves absolutely jobless in the time after the 10th board exam, Jay was looking for something interesting to do and decided to try his hand at C. However it wasn't until the second semester in college that he started pursuing coding seriously.Talking about how he ventured into this world he says "It was my first interaction with Linux and I was very excited to explore all these things. It felt as if the terminal was giving me some power over the computer and I could do whatever I want." Tweaking and fiddling with the code to create something new enthralled him. A journey had begun.

Having done a digital image processing course, Jay had some basic image processing knowledge, and was looking for some long term project in similar fields. The Google Summer of Code presented itself as the perfect opportunity. At that point python was the only programming language he knew well. Simple CV is a computer vision based framework written in python. Seeing the dots connect, Jay gave it his all and made it through

On selection for GSoC, he worked on a state of the art tracking system called TLD, a Phd thesis paper by Dr. Zdenek Kalal. His project was to port the tracking system to Python and use that technology to segregate unwanted items from the conveyor belt in an industry. The project however involved a lot of advanced machine learning algorithms, and having limited knowledge of these, he couldn't complete the said project.

He and his project instructor both agreed that the project was a long shot. Having cleared the mid term evaluation he implemented 3 advanced tracking algorithms for OpenCV for the rest of the coding period.

"I got to learn a lot from the project. I realized that there are some things that you can not do at present, but that doesn't mean that you should not try" says Jay in retrospect. Even after the internship period Jay continued contributing to SimpleCV and infact contributed a lot more during the subsequent period than he had during the actual internship.No wonder, the SimpleCV guys selected him once more, for GSoC 2013.

This year he is going to work on SimpleCV 2.0 which is going to be highly optimized, less bulky and more speed efficient..His part will be to develop 2D shape matching and detection features with some other minor features.


If two GSoCs weren't impressive enough Jay also was also selected for the design innovation workshop by MIT Media Labs. “What I loved about the workshop was that it wasn’t just an ordinary tech workshop. It focussed a lot on the value of great design.” says Jay.He along with his teammates Aditi(from IIT-M) and Nikita(from IIIT-H) developed a basic prototype of a pen for blind and autistic people which provide haptic feedback. The project required a webcam attached to a computer, a pen and a small vibrator motor and tracked the movement of the pen using the camera and sent feedback using the motor. They had built a working prototype by the end of the event.


He is currently working as an intern at Ducere, for whom he had initially worked remotely earlier. At that time he was working on an obstacle detection technique which they have implemented in their prototype shoes for blind people - 'Le Chal'.The shoes have sonar and couple of small cameras which detect obstacle in front of the blind people and hence indicating them to stop and turn in other direction. He is currently working with them as a full time intern and is working on 3D gesture recognition using Kinect.

When Jay isn’t behind his laptop coding his way to glory, he loves to play Ultimate Frisbee. He is also an ardent follower of Douglas Adams' philosophy and his books. Apart from that, he enjoys listening to 60s and 70s rock music- Led Zeppelin and The Doors being his favorite.

Currently his interests tend towards scientific computing.He intends to make a library for Computational Fluid Dynamics in Python and also some FOSS(free and open source software) to replace MATLAB. The advent of products like Google glass and Microsoft kinect are sure to give great impetus to the field of computer vision.With Jim Morrison's immortal words “ I am the lizard king, I can do anything” as his personal mantra, and computer vision that is both his forte and his passion,Jay shows great promise in the years to come. We, at YourStory wish him luck for the journey ahead.

Website :Jay Rambhia

If you think you deserve to feature in this column or know anyone who does, you can drop a mail at ganesh[at]yourstory[dot]in

Get access to select LIVE keynotes and exhibits at TechSparks 2020. In the 11th edition of TechSparks, we bring you best from the startup world to help you scale & succeed. Join now! #TechSparksFromHome

    Clap Icon0 Shares
    • +0
      Clap Icon
    Share on
    close
    Clap Icon0 Shares
    • +0
      Clap Icon
    Share on
    close
    Share on
    close