Current Research:
Mult-modal learning: Multi-modal learning is a type of machine learning where a model learns from more than one kind of data, or modality, at the same time.
Object counting in Images and Video: The object counting problem is the task of determining how many objects of a particular type are present in an image, video, or scene.
Past Research Projects:
Data Driven Material Discovery Center: This is an NSF funded collaborative project with University of South Dakota, South Dakota State Mines and Technology, and Montana State University. I lead the development of Biofilms Data and Information Discovery system that will integrate metadata from accessible materials and biofilms data sources, employs it to process natural language processing (NLP) queries from users to predict biofilm phenotype on a material.
Study of Eye Disease Progression Using Retinal Images: This is a collaborative project with ophthalmologists from the Byers Eye Institute, Stanford University. The goal of the project is to develop scalable machine learning algorithms analyze retinal images to understand the incidence and progression of ocular toxoplasmosis and age-related macular degeneration. My role is to design and develop the data management and machine learning tools needed for image analyses.
Biofilm Priority Questions Project: I am collaborating with researchers from University of
South Dakota, Center for Biofilm Engineering, U.S, COST AMiCI Consortium, Finland, ESCMID Study Group for Biofilms (ESGB), Belgium, National Biofilms Innovation Centre (NBIC), U.K, and Centre for Environmental Life Sciences Engineering (SCELSE), Singapore in Priority Questions Exercise for microbial biofilms. My role is to lead the development of text mining tools needed to support the project.