Avatar

Naman Biyani

3rd year Undergraduate

Indian Institute of Technology Kanpur

About Me

I’m a Prefinal year UG student at Indian Institute of Technology, Kanpur majoring in Computer Science and Engineering, broadly interested in the areas of Deep Learning and its applications. I have explored the fields of Computer Vision and Natural Language Processing wherein I was fascinated by Deep Generative Models, Recommender systems, Variational Autoencoders and the advancement in Object detection using Deep Learning methods. I had some exposure to Probabilistic Machine Learning in Summer 2019, learning about Bayesian regression methods, applications of Variational Inference and MCMC algorithms and I am interested in learning more about them. I have been working on video generation and prediction using Stochastic and adversarial ways. I am also studying more about multi-domain translational GANs,cool generative architectures(like StyleGAN), Mask R-CNNs, VQ-VAEs and text-generative network architectures.I also had some exposure in Reinforcement learning where I studied Monte-Carlo learning,Temporal learning and Q-learning and trained 2 Atari games using Deep-Q learning and A3C methods.

Currently, I am working as a Research Assisstant(work from home) under Prof. Yogesh Rawat in Computer Vision Lab of University of Central Florida on the topic of Conditional Video synthesis.I am also part of a Probabilistic Machine Learning reading group of Prof. Piyush Rai where we weekly discuss few of the latest research papers on Probabilistic Machine learning.

Interests

  • Video Synthesis
  • Deep Generative Models
  • Object Detection
  • Novel view synthesis
  • Representation Learning
  • Probabilistic Machine Learning

Education

  • BTech in Computer Science and Engineering, 2018-Present

    Indian Institute of Technology Kanpur, India

Experience

 
 
 
 
 

Research Assistant(Work from home)

Computer Vision Lab, University of Central Florida

April 2020 – Present Florida, USA

Projects

*

Independent research in Deep Learning Applications

Working on making new adversarial ways for video generation and prediction. Exploring topics such as StyleGANs, NLP-GANs, MoCoGANs, SeqGANs, VQ-VAEs and Mask-RCNNs and implementing them .

Probabilistic Machine Learning

Deep-dived and implemented Batch and Online EM methods,Blackbox VI, Reparamaterization trick in Variational Inference, Stochastic VI and Recommender systems using Bayesian Matrix factorization.

A Study in GANs

Studied Convolutional nueral networks and Generative adversarial networks in-depth . Implemented and played with DCGAN, class conditioned GANs like ACGAN,CGAN,InfoGAN and Style Transfer GANs like DiscoGAN, CycleGAN and StarGAN . Implemented dataloader and progress bar feature in TorchGAN(research framework to train GANs) and tried implementing a YAML parser to train GANs automatically .

Reinforcement Learning in Atari Games

Studied RL through David Silver's lectures and solved Dennybritz's excercises . Implemented DQN and A3C algorithms in Pong and Breakout using Pytorch and OpenAI gym.

Haskell Scrabble Solver

Deep-dived into functional programming and made a Scrabble Solver in Haskell ( A Two Player version and a PlayWIthComputer version) which used Lexicograhical Search, Regex-type functions(wriiten from scratch) and Quick Sort as the major algorithms

Voting App using Blockchains

Made a voting app(for Microsoft Codefundo competition 2019) which used private blockchains(using Microsoft Azure blockchain services) for security.

Framework to determine resource allocation to eradicate Polio

Worked with Rotary club as a Data science team member to make a framework for resource allocation to eradicate Polio if it reoccurs. Can be geralized to other epidemics...

Contact