Vaswani's recent research has focused on provably correct and practically useful online (recursive) algorithms for various structured (big) data recovery problems. She has worked on (a) dynamic compressive sensing (CS), (b) dynamic robust principal component analysis (RPCA), and most recently on (c) Phaseless PCA and Subspace Tracking (structured phase retrieval). Online algorithms are needed for real-time applications, and even for offline applications, they are typically faster and need less storage compared to batch techniques. Most importantly, her work on these problems has shown that online solutions provide a natural way to exploit temporal dependencies in a dataset, without increasing algorithm complexity; and that exploiting such dynamics provably results in either reduced sample complexity (in case of dynamic CS) or improved outlier tolerance (in case of dynamic RPCA). The former implies proportionally reduced acquisition time for applications such as MRI where data is acquired one sample at a time. The latter implies increased robustness to outliers such as large-sized or slow changing foreground occlusions in videos. All theoretical claims are backed up by extensive experimental evaluations for various video analytics applications and medical imaging applications. In the past she has also worked on particle filtering (sequential Monte Carlo) algorithms, and in computer vision.
Professor of Electrical and Computer Engineering
Professor of Mathematics
Area of Expertise:
Intersection of statistical machine learning / data science
and signal processing
Ph.D. in 2004 from the University of Maryland, College Park
B.Tech. from Indian Institute of Technology (IIT-Delhi) in India in 1999