This talk is about two of my recent research works.The first work concerns a reconstruction problem in multi-dimensional magnetic resonance imaging (specifically,high-angular-resolution diffusion imaging), using acquisition patterns with sub-Nyquist sampling rates. Our contributions involve a compressed-sensing formulation that proposes rotation-invariant dictionaries and combines them with over complete wavelet frames for signal modeling.The second work concerns analysis of shapes where each shape belongs to one of multiple groups within a population. Our contributions involve a hierarchical graphical model for shapes and an efficient sampling scheme in shape space. We show applications in hypothesis testing and classification.
Suyash Awate is a research assistant professor in the School of Computing and a faculty member in the Scientific Computing and Imaging (SCI) Institute at the University of Utah.He previously worked at the University of Pennsylvania and Siemens Corporate Research.His research interests include medical image processing and computer vision, relying on statistical inference and pattern recognition.