About me

Biography

I am a Computer Science Masters student at McGill University. I'm currently working in the Shape Analysis group under the guidance of Prof. Kaleem Siddiqi . My areas of interest are computer vision, computational perception and machine learning. I am intrigued by the role of computer vision in biological systems and my current work is related to modeling chirality in vaterite crystals found in nature. Here are a few of my research interests, past internships and projects:

Internships

I did my summer internship recently at NEC Laboratories America ,Princeton, on the application of deep learning in optical network and sensing. I was a student research assistant at Karlsruhe Institute of Technology in the summer of 2016 where my work pertained to the development of a framework for multi-scale material simulation. I also did an internship at Indian Institute of Technology, Delhi, on content management system during my undergrad.

Projects at McGill

  • Modeling chirality: simulating left/right handedness in vaterite crystals
  • Rick and Morty: Implemented key frame animation using hierarchy of transformations in OpenGL
  • Image captioning using deep learning: explored niceties of architectures like LSTM, Gated Recurrent Unit, Bidirectional LSTM to provide semantic descriptions to images and results
  • ICLR Reproducibility challenge 2018: implemented the findings of a paper called "What are image captions made of?" which focused on improvement of image captioning systems by using different image representations
  • Language classification: analyzed and classified short conversations according to languages with Naive-Bayes and K-Nearest Neighbors models for language classification
  • Implemented stochastic programming to find optimal path to maximize the dirt picked up by Roomba (a self driving cleaning device) across different dirt distribution scenarios

Extra-curricular activities

I am a Kathak, contemporary dance and Fine Arts enthusiast and enjoy doing creative activities in my spare time!