Projects

Here is a non-exhaustive outline of my non-research projects. My research projects can be found over here.

2018

  • Solar and Vibration Powered Portable Charger - Designed and created a charger to harvest the energy of renewable sources like solar and vibration energy and use it to charge a mobile device. Energy from vibrations is harvested using the principles of electromagnetic induction and piezoelectricity. The project involved building a Buck-Boost converter to regulate the energy accumulated using solar and vibrational energy.

  • Bellatrix - We project a game on a screen, and a camera records the user interaction with the screen. The key challenge here was accurately determining a foreign object in the camera recording and deciding if it touches the screen. We use the difference of the image being projected and that being recorded to localize the foreign object and then track its speed to determine the hit location. We presented this at a hackathon organized by Yahoo! Japan and won the “Best Tech Prize” at the event among 100+ teams.

2017

  • Localising Fiducials for Neuroregistration - Participated in a pan-IIT competition, and developed an algorithm to estimate the location of fiducials (adhesive markers) affixed onto the skull from a series of MRI images, provided in the DICOM format by using mean-shift segmentation and iterative closest point matching. Developed techniques to identify surface voxels at the surface with template fiducial depth maps. Achieved the second position and received a silver medal from Bhabha Atomic Research Centre Mumbai. They plan to use the algorithms developed by the top three teams for indigenous technology under the Make in India initiative.

  • Face Swapping - Developed an algorithm to swap the faces of people with standard faces to protect the privacy of individuals. Our algorithm identifies and uses the most suitable faces using a selection process, extracts the face using optimal seam search, and blends using linear RGB channel scaling and Poisson Editing.

  • Direct-sequence spread spectrum - Implemented a DSSS spread spectrum modulation technique in GNU Radio and demonstrated its resistance to interference by showing that only the intended receiver is able to decrypt the sent message.

  • Google Summer of Code - Selected for the Prestigious Google Summer of Code 2017 which had an acceptance rate of 7% for Oppia a globally reputed online interactive learning platform. Designed efficient algorithms, scalable database models and computed complex statistical information using parallel programming architectures like Map-Reduce to track the progress of millions of learners on the site in real time. Collaborated with multiple teams, including the design team to create a user-friendly dashboard helping learners access this information and gain relevant insights about their progress.

2016

  • Mood Indigo App - Lead developer of the Official Mood Indigo 2016 App. The app amassed more than 5000 downloads and has a rating of 4.8 stars.

© 2018. All rights reserved.

Powered by Hydejack v7.5.0