Events Calendar

Geometry as a Prior in Signal Processing

Speaker: Yuejie Chi, Ph.D., Department of Electrical Engineering, Princeton University

Talk Title: Geometry as a Prior in Signal Processing

Abstract: My theme is that geometry provides a compressive representation of the worlds that surround us. I will present examples of waveform design in radar and communications where geometric structure is used to probe the physical world more precisely. I will present applications to mobile healthcare where the focus is preserving diagnostic content in wireless Electrocardiogram monitoring using body area networks. When signals of interest live in a low dimensional subspace that evolves over time I will derive a new algorithm (PETRELS) that is able to track the subspace from incomplete data and then demonstrate improvements to Direction of Arrival Estimation in array processing. Finally I will address mismatch between the actual and the assumed geometry which limits the opportunity to simplify signal representation and reduce the complexity of signal processing.

Biography: Yuejie Chi is a Ph.D. candidate in Electrical Engineering at Princeton University, advised by Professor Robert Calderbank. She received her B.Eng. in Electrical Engineering from Tsinghua University 2007 and her M.A. from Princeton University in 2009. She has interned with the Machine Learning Group at Mitsubishi Electric Research Lab in Cambridge, MA in 2011 and with the Mobile Health Group at Qualcomm Inc. in San Diego, CA in 2010, where she won a Roberto Padovani Scholarship. Her research interests include statistical signal processing, high-dimensional data analysis, machine learning and her work is inspired by applications to communications, networks, wireless healthcare, sensing and image processing.



Host: Professor C.C. Jay Kuo

Location: Hughes Aircraft Electrical Engineering Center (EEB) - EEB

Audiences: Everyone Is Invited

Contact: Talyia Veal