I am a senior undegraduate student in Computer Science at the Indian Institute of Technology Kanpur. Recently I visited Prof. Vincent Ng at the University of Texas at Dallas, where I worked on event coreference resolution. Previously, I have worked with Prof. Ragunath Tewari on probabilistic polynomials in circuit complexity and with Prof. Amitabha Mukerjee on generative image models for path planning on the CRS Robot Arm.
When I’m not occupied with my academic endeavours, I like to hone my speed solving timings for the Rubik’s Cube and its other variants, such as the 4 x 4, one-handed 3 x 3, Pyraminx and many more! My official World Cube Association record for fastest single solve of the 3 x 3 is 15.33 seconds.
You can find my resume here.
B. Tech in Computer Science, 2017
Please see my resume for a list of all my projects
Implemented three variations of the Deep Recurrent Attention Writer (DRAW) model for incorporating convolutional features and supervision to the existing encoder-decoder archirecture.
Implemented an RNN model for identifying grammatical errors using the NUCLE corpus of the CoNLL-2013 shared task, followed by testing on the NIPS 2015 publications dataset.
Studied the application of the probabilistic polynomial method by Ryan Williams in the All Pairs Shortest Path and Boolean Orthogonal Detection problem.
Implemented an image colorization model which learns to color grey scale images pixel by pixel using local patch luminance based inference. This idea was inspired by the super-resolution model of Yang et. al. in "Fast Direct Super-resolution by Simple Functions"
Tutor - I conducted weekly tutorial classes, designed and evaluated the exam and lab questions for the following courses:
ESC101: Fundamentals of Computer Science (2016-17 I), by Prof. Vinay Namboodiri
ESC101: Fundamentals of Computer Science (2016-17 II), by Prof. Rajat Mittal