Biography

Welcome to my academic homepage!

I am a first-year Computer Science graduate student at the University of Maryland, College Park. My research interests lie in the intersection of Computer Vision, Computer Graphics and Deep Learning with an application domain of Augmented and Virtual Reality.

Previously, I was a researcher at Innovation Labs, Tata Consultancy Services collaborating with Dr. Lovekesh Vig and Ms. Ramya Hebbalaguppe. Prior to this, I completed my undergrad with Computer Science and Engineering major at Indraprastha Institute of Information Technology Delhi (IIITD). At IIITD, I was fortunate to work with Prof. Saket Anand and Prof. Ojaswa Sharma on interesting problems of scene reconstruction and procedural modelling.

Longer version of my resume is available here.

Interests
  • Deep Neural Network Optimization
  • 3D Scene Understanding
  • Egocentric Video Analysis
  • Situated Visualization in AR/VR
Education

  • M.S. in Computer Science, 2021-2023

    University of Maryland, College Park

  • B.Tech in Computer Science and Engineering, 2017

    Indraprastha Institute of Information Technology Delhi

Recent Updates

  • 1st Aug 2021: My co-authored work “Rethinking Common Assumptions to Mitigate Racial Bias in Face Recognition Datasets” accepted at ICCV (HTCV ‘21) as a oral paper. My first publication as a UMD student!
     

  • 25th May 2021: I have started my grad school at University of Maryland, College Park.  

  • 13th Apr 2021: My co-authored work “Cross-Domain Multi-task Learning for Object Detection and Saliency Estimation” accepted at CVPR (CLVISION ‘21) as a poster.
     

  • 14th Feb 2020: Received TCS Citation Award for the second time for outstanding contributions to the organization through publications.
     

  • 20th Jul 2020: Invited as a reviewer for IEEE Transactions on Circuits and Systems for Video Technology (TCSVT, impact factor: 3.599). My first journal reviewing experience!
     

  • 21st Mar 2020: TreasAR Hunt - an AR based treasure hunt - organized for Re.Fresh 2020 at TCS Innovation Labs New Delhi. The source code is now available. Check it out!
     

For older news visit here.

Experience

 
 
 
 
 
Researcher, IT Analyst
TCS Innovation Lab
Aug 2017 – Present New Delhi

Responsibilities include:

  • Research - Computer Vision, Deep Learning and Artificial Intelligence
  • Developing Augmented Reality prototypes
  • Prototyping and Deploying
 
 
 
 
 
Teaching Assistant
GPU Computing at IIIT-Delhi
Jan 2017 – May 2017 New Delhi
Designing course curriculum, CUDA assignments and projects for the course of GPU Computing at Graduate and Post graduate level at IIIT-Delhi.
 
 
 
 
 
Software Development Intern
HEICO, India
Jan 2017 – Mar 2017 New Delhi
  • Fine-tuned control of Fatigue Testing Machine(FTM) used for evaluating material strength and fatigue visualization.
  • Worked with .NET technology for developing data analysis tool for PID controller data of FTM.
 
 
 
 
 
Research Intern
TCS Innovation Lab
May 2016 – Aug 2016 New Delhi
  • Worked on Augmented Reality with Google Cardboard for Android Platform.
  • Development of hand gesture interaction techniques for Google Cardboard for Augmented Reality.
  • UI development and text rendering for Google Cardboard application.
 
 
 
 
 
Teaching Assistant
Data Structures and Algorithms at IIIT-Delhi
Jun 2015 – Aug 2015 New Delhi
Summer refresher module and orientation for the incoming M.Tech batch of 2015.

Projects

Deep Neural Network Compression

Compressing deep neural network models for memory constrained devices
Deep Neural Network Compression

Reverse VooDoo

Automating animation through motion transfer from real scenes.
Reverse VooDoo

Non Intrusive Overlay Placement in User Interfaces

A non-intrusive, relevant and temporally coherent overlay placement technique
Video
Non Intrusive Overlay Placement in User Interfaces

Draw In Air

A hand gesture based in-air touch-less interaction technique for AR applications
Video
Draw In Air

Robust 3D Reconstruction of Indoor Scenes using Deep Learning

Developing an end to end pipeline for automatic generation of 3D models of indoor scenes.
PDF Dataset Video
Robust 3D Reconstruction of Indoor Scenes using Deep Learning

Distributed Fault Tolerant Multi-Robot Area Coverage Under Limited Communication Ranges

Analysed Multi-Agent area coverage problem for surveillance purposes.
Video
Distributed Fault Tolerant Multi-Robot Area Coverage Under Limited Communication Ranges

Modelling Vegetation with L-Systems Using an Image

Procedural modelling of vegetation using context-free grammar
PDF Code
Modelling Vegetation with L-Systems Using an Image

Vision Based Outdoor Localization

Outdoor location estimation in IIIT-Delhi campus through computer vision
PDF Slides
Vision Based Outdoor Localization

Sketch23D

A sketching interface for rapid designing of freeform 3D models from 2D sketches
Code Poster
Sketch23D

Virtual Campus Project

Creating a virtual walk-through of IIIT-Delhi campus
Project
Virtual Campus Project

Recent Publications

Quickly discover relevant content by filtering publications.
M. Gwilliam, S. Hegde, L Tinubu, A Hanson (2021). Rethinking Common Assumptions to Mitigate Racial Bias in Face Recognition Datasets. In HTCV, ICCV.

Preprint PDF Code

A. Khattar, S. Hegde, R. Hebbalaguppe (2021). Cross-Domain Multi-task Learning for Object Detection and Saliency Estimation. In CLVISION, CVPR.

PDF

S. Yalburgi, T. Dash, R. Hebbalaguppe, S. Hegde, A. Srinivasan (2020). An Empirical Study of Iterative Knowledge Distillation for Neural Network Compression. In ESANN.

PDF

S. Hegde, R. Prasad, R. Hebbalaguppe, V. Kumar (2020). Variational Student: Learning Compact and Sparser Networks in Knowledge Distillation Framework. In ICASSP (Oral).

Preprint PDF

S. Hegde, J. Maurya, A. Kalkar, R. Hebbalaguppe (2020). SmartOverlays: A Visual Saliency Driven Label Placement for Intelligent Human-Computer Interfaces. In WACV.

PDF Project Video

See all publications

Contact