Research: Ideas & Work

This page is to curate my journey in research. While this page would be very similar to my other official profile pages, it is different too -- I will be using this page to reflect upon some of my current work and its implications.

I lead the Graphics-Visualization-Computing Lab (GVCL) at IIIT Bangalore (IIIT-B), where I am a faculty member. I am affiliated with the E-Heath Research Center (EHRC) at IIIT-B. My publications and other artifacts are regularly updated on the lab webpage. Before that, I had worked on research at TACC at the U.T. Austin, IDAV/CIPIC at the U. C. Davis, and ERC at the Mississippi State University.

My current research interests involve the use of novel ensemble methods to improve the certainty of analysis of data, and data representations such as networks and matrices. I am particularly interested in applications with complex data with inherent uncertainty, with a focus on spatiotemporal (spatial and/or time series), and count data.  Thus, I find scientific, engineering, and population studies datasets fascinating. I currently identify my research with topics of environmental perception and human interactions with environments (both technological and natural).

The research questions are in quantifying the uncertainty in the data science workflows, preparing appropriate representations of the datasets, and designing and implementing their corresponding data science workflows for effective visualization, exploration, and knowledge discovery. 

A few of the data science workflows I have worked on are for visualization and analysis of:

• geoscience/earth observations (LiDAR point clouds, SAR images, multispectral images, cartographic maps);

• chart images (bar, line, pie, scatter charts, cartographic maps); 

• environmental data (floods, air pollution);

• human-computer interaction (HCI) data (gaze and mouse tracking);

• population (count) data (national health surveys in India, COVID case logs in Singapore and S. Korea, migration flow);

• biomedical data (brain functional connectome, multi-omics cancer data);

• traffic data (LiDAR point clouds, monocular images).  

For my research in my stints before IIIT-B/GVCL/EHRC, I have worked on 

• scientific visualization of gridded data (isosurface extraction from scalar fields with hexahedral cells in 3D volumes, topological analysis of second-order tensor fields in 2D data), and building a visualization tool for tabular data (specifically of the California river system and water reservoirs), during 2002-09; 

• feature extraction from vorticity maps (images) from fluid flow data using wavelets, and feature ranking, during 2000-02;

• polygonal elements in FEM (structural mechanics), during1999-2000.  

[This website is under construction.]