Namrata Deka

Namrata Deka

Graduate student

The University of British Columbia


Hello! I am a master’s student at UBC, Vancouver working on deep kernel-based representation learning for computer vision applications under the supervision of Dr. D.J. Sutherland. I am also working with Dr. Amir Zamir in the Visual Intelligence and Learning Lab at EPFL on learning sparse representations of articulated objects from videos and optical flow to create robotic object models.

Prior to starting grad school, I was a research fellow at Wadhwani AI for two years where I was involved in the research and development of an AI-powered visual screening tool to detect low birth-weight (<2.5kg) babies in rural India where primary healthcare is inadequate. I have also interned at Microsot Research India's Applied Sciences group on prototyping machine learning techniques to identify highly imbalanced P.I.I. attributes from massive customer databases for the EU GDPR mandate.


  • Artificial Intelligence
  • Causal Modelling
  • Computer Vision
  • Reinforcement Learning
  • Wildlife Conservation
  • Climate Change
  • Sustainable Development


  • MSc in Computer Science, 2022

    The University of British Columbia, Vancouver

  • BTech (Hons) in Computer Science & Engineering, 2017

    Indraprastha Institute of Information Technology, New Delhi (IIITD)



Summer@EPFL Fellow

Visual Intelligence and Learning Lab (VILAB), EPFL

June 2021 – Aug 2021 Lausanne, Switzerland
Sparse representation learning of articulated objects from optical flow.

Graduate Research Assistant

University of British Columbia

Sep 2020 – Present Vancouver, BC
Deep kernel-methods for representation learning.

Graduate Teaching Assistant

University of British Columbia

Jan 2021 – Apr 2021 Vancouver, BC
CPSE 422: Intelligent Systems

Research Fellow

Wadhwani Institute for Artificial Intelligence

Aug 2018 – Aug 2020 Mumbai
AI for Social Good: Screening low birth-weight babies in rural India using 3D vision.

Research Intern

Microsoft Research

Jan 2018 – Jul 2018 Bengaluru
Classification of higly imabalanced massive datasets.