Biography

I am an incoming PhD student in Cognition & Perception at New York University. I recently completed my undergraduate degree in Electronics and Communication Engineering from BITS Pilani, India. I am primarily interested in developing computational models of human perception.

Previously, I have worked on projects spanning computer vision, reinforcement learning and cognitive neuroscience in collaboration with Harvard University, CCBR, Max Planck Institute for Psycholinguistics, Human Brain Project and INCF

Interests
  • Deep learning
  • Vision science
Education
  • Ph.D. in Cognition & Perception, 2021-2026

    New York University

  • B.E. in Electronics & Communication Engineering, 2017-2021

    BITS Pilani

Experience

 
 
 
 
 
Research Intern
Jun 2020 – Present
Human-level learning in Atari-like video games using a model-based reinforcement learning approach inspired by theories of human cognition.
 
 
 
 
 
Research Intern - Undergraduate Thesis
Jun 2020 – Dec 2020
Developed a deep learning based cellular segmentation model for gigapixel resolution neuroanatomical images.
 
 
 
 
 
Research Assistant
Aug 2019 – Dec 2020 Goa, India
  1. Deep learning to understand how spoken words are visually represented in the brain. Collected EEG data from human subjects listening to spoken audio of numerical digits, and employed deep generative models to construct images of digits purely from EEG signals.
  2. Review paper on the neural and psychological basis for reinforcement learning algorithms.
 
 
 
 
 
Research Assistant
Aug 2019 – Jun 2020 Goa, India
Using a SpiNNaker spiking neural network model to validate experimental results relating to synchrony, periodicity and luminance response of the Lateral Geniculate Nucleus (LGN) in a mouse brain.
 
 
 
 
 
Google Summer of Code Intern
May 2019 – Aug 2019
Developed TensorGeNN, an open source Python library to convert trained deep neural network models to spiking neural networks with minimal losses in performance. Library was benchmarked on MNIST and CIFAR-10 datasets.