Kundan Kumar
Email: kundankumar2510-AT-gmail.com

I co-founded Lyrebird-AI. We are hiring awesome software developers. Please contact me if you are interested.

I am a PhD Student at MILA advised by Aaron Courville and Yoshua Bengio. I graduated with B. Tech. and M. Tech. degrees in Computer Science and Engineering from IIT Kanpur.

I visited MILA, University of Montreal as an intern in 2016. I was advised by Prof. Vinay P. Namboodiri and Prof. Yoshua Bengio for my masters thesis.

I visited Computer Vision lab, UMASS, Amherst in summers 2015 where I worked with Prof. Subhransu Maji

Github | LinkedIn

  • Two papers (sampleRNN and pixelVAE) are on arXiv.
  • Paper explaining texture generation networks accepted at ICASSP, 2017.
  • My team won Xerox Research Innovation Challenge held in Decemeber 2015. The competition blog post lies here.

[NEW] SampleRNN: An Unconditional End-to-End Neural Audio Generation Model
Soroush Mehri, Kundan Kumar, Ishaan Gulrajani, Rithesh Kumar, Shubham Jain, Jose Sotelo, Aaron Courville, Yoshua Bengio
Accepted at ICLR, 2017


[NEW] PixelVAE: A Latent Variable Model for Natural Images
Ishaan Gulrajani, Kundan Kumar, Faruk Ahmed, Adrien Ali Taiga, Francesco Visin, David Vazquez, Aaron Courville
Accepted at ICLR, 2017


[NEW] On Random Weights for Texture Generation in One Layer Neural Networks
Mihir Mongia, Kundan Kumar, Akram Erraqabi, Yoshua Bengio
Accepted at ICASSP, 2017

Template: this, this, this, this and this