Cleared GATE-2016 with a rank of 810/152318(99.47 percentile).
Coursework: Probabilistic Mobile Robotics, Signal Processing, Image Processing, Mathematical methods and Modelling.
Joined Team IITK-TCS for Amazon Robotics Challenge 2017. Roles included programming low and high-level image processing, load balancing, and reducing latency. Won 3rd, 4th and 5th prizes in the event in Robocup 2017, Nagoya, Japan.
Worked on extracting depth information from Light-field data using a simulated annealing algorithm.
Used a Light-field digital refocusing algorithm to patent a novel orthographic imaging algorithm.
Developed a portable version of a device capable of counting 200-micron steel sheets with an accuracy of 99.6% using the algorithm mentioned above.
Developed physical hardware for a Light-Field capture system for capturing images across a plane.
Demonstrated orthographic capture's generalized ability using the patented algorithm to count bundled straws with occlusions. The technique is one of the ways to count heavy objects like steel rebars/pipes.