Shruthi Kubatur's homepage

"An optimist will tell you the glass is half-full; the pessimist, half-empty; and the engineer will tell you the glass is twice the size it needs to be." - Oscar Wilde

"Forget the one glass and the bunch of philosophies for a while; come back to me when you have a large enough sample space to make sensible, non-cliched, decisions" - Shruthi Kubatur


I am a doctoral student in the School of Electrical and Computer Engineering at Purdue University, West Lafayette, USA.

I currently work on advanced simulation techniques for rare-events in natural systems.

I am primarily interested in:

  • Stochastic modeling
  • Advanced simulation methods
  • Image Processing
  • Model-based imaging

Ouline of current research [link to my papers]

Stochastic models of images are very useful for applications such as segmentation, deblurring, and reconstruction. Sometimes it is important to be able to simulate, or draw samples from, a stochastic image model. For example, simulation can be used as an optimization tool for segmenting, deblurring, or reconstructing an image. Also, simulation of images that characterize a system can be helpful in understanding the system, by allowing virtual exploration of models of the system instead of expensive and time-intensive physical experimentation. There are of course many Markov chain Monte Carlo (MCMC) methods for drawing samples from the ubiquitous Markov random field (MRF) image model. However, these methods draw sample images that represent typical cases of the model. To sample images that occur with low probability, which represent rare events, a prohibitive number of Monte Carlo samples would need to be drawn using traditional MCMC.

In this research work, we turn to large deviations theory and importance sampling to propose a rare-event simulation method for MRFs. We then use an impactful problem from materials science to demonstrate the application of our method. More specifically, we look at the phenomenon of abnormal grain growth in polycrystalline materials. With our proposed method, we consistently generate images containing abnormal grain growth, though this is a very challenging problem for standard Monte Carlo simulation methods. Importantly, our method can be used to simulate rare events in a broad class of imaging applications, namely those that use an MRF model.

In a similar spirit, we are now extending the Gibbs distribution to model microstructures of certain important classes of alloys. Further, we are successfully simulating abnormal events that occur in those microstructures. More to come; stay tuned!

My previous research projects:

Purdue Electronic Imaging Systems Laboratory
Worked on image processing problems in developing stochastic printer models and integrating it with Direct Binary Search (DBS) algorithm for optimized digital halftoning.
Prior to this, I worked on clustered DBS algorithms. Performed major code-drop to HP Labs, Israel.

Purdue Robot Vision Laboratory
(1) Computer vision and detection algorithms (mainly Viola Jones/ AdaBoost).
(2) Information retrieval in software repositories. (An Infosys Technologies project.)

University of Windsor, Canada
Unconstrained handwriting recognition via artificial neural networks.
Accuracies of up to 97.6% were achieved, and several novel features were introduced. I built the world’s first Devanagari handwriting database with layered demographic texture. Worked on novel methods of feature extraction and recognition.

V.T.U., India
Built a linguistic translator using optical character recognition and transliteration. Introduced a novel method of grammar number that helps in grammar correction of transliterated and translated sentences. Also worked on creating a look-up table for quick and efficient transliteration.

hit counter website
hit counter website