Shamak Dutta

s7dutta at uwaterloo dot ca

I am a PhD student in Electrical and Computer Engineering at the University of Waterloo, where I am advised by Stephen Smith. My research interests are in planning, decision-making, and learning methods for robotics.

I did my master's in Systems Design Engineering where I was advised by Bryan Tripp and Graham Taylor. My thesis was on the analysis of one of many differences between biological and artificial neural networks. I completed my bachelor's in Electrical and Computer Engineering where I worked with Hamid Tizhoosh on image retrieval, Dana Kulic on behaviour cloning for human motion, and Stephen Smith on the set traveling salesman problem.

I've collaborated with Shunta Saito and Masaki Saito at Preferred Networks in Tokyo, Japan on high resolution video prediction using deep learning. Prior to that, I interned at Waymo UK (Latent Logic) collaborating with Joao Messias and Shimon Whiteson on 2D to 3D pose estimation. I've also had the opportunity to work with Bing Yin and Erick Cantu-Paz on improving search ranking using machine learning at Amazon Search, Palo Alto.

CV  /  Google Scholar  /  GitHub

Research
Convolutional Neural Networks Regularized by Correlated Noise
Shamak Dutta, Bryan Tripp, Graham Taylor
Conference on Computer and Robot Vision, 2018
bibtex

We show that adding spatially and tuning dependent correlated noise to the activities of units in a neural network helps in regularization and image classification under occlusion.

Barcodes for Medical Image Retrieval Using Autoencoded Radon Transform
H.R. Tizhoosh, Christopher Mitcheltree, Shujin Zu, Shamak Dutta
International Conference on Pattern Recognition, 2016
bibtex

We show that barcodes generated by thresholding the learned representations of autoencoders on the Radon transform of images can be used in medical image search and retrieval.

Teaching Assistantships

ECE 406: Algorithm Design and Analysis taught by Stephen Smith in Winter 2020.


credits