【2018學術報告11】On Deep Learning Based Indoor Localization
Title: On Deep Learning Based Indoor Localization
Speaker: Shiwen Mao，Professor, Auburn University
Time: 09:30 a.m., 24th, Oct. (Wed.)
Place: 1-415, FIT Building
Organizer: Research Institute of Information Technology (RIIT), Tsinghua University
Shiwen Mao received his Bachelor's and Master's degrees, both in Electronic Engineering, from Tsinghua University, Beijing, P.R. China in 1994 and 1997, respectively. He also received a Bachelor's degree in Business Management from Tsinghua University in 1994 and a Master's degree in Systems Engineering from Polytechnic University (now NYU Tandon School of Engineering), Brooklyn, NY, in 2000. He received his Ph.D. in Electrical and Computer Engineering from Polytechnic University in 2004. He is the Samuel Ginn Distinguished Professor and Director of the Wireless Engineering Research and Education Center (WEREC) at Auburn University, Auburn, AL. His research interests include wireless networks, multimedia communications, and smart grid. He is a Distinguished Speaker of the IEEE Vehicular Technology Society (VTS). He is on the Editorial Board of IEEE Transactions on Mobile Computing, IEEE Transactions on Multimedia, IEEE Internet of Things Journal, IEEE Multimedia, ACM GetMobile, among others. He received the Creative Research & Scholarship Award from Auburn University in 2018. He received the 2017 IEEE ComSoc ITC Outstanding Service Award, the 2015 IEEE ComSoc TC-CSR Distinguished Service Award, the 2013 IEEE ComSoc MMTC Outstanding Leadership Award, and the NSF CAREER Award in 2010. He is a co-recipient of the IEEE ComSoc MMTC 2017 Best Conference Paper Award, Best Paper Awards from IEEE GLOBECOM 2016 & 2015, IEEE WCNC 2015, and IEEE ICC 2013, and the 2004 IEEE Communications Society Leonard G. Abraham Prize in the Field of Communications Systems.
With the fast growing demand of location-based services in indoor environments, indoor positioning based on fingerprinting has attracted considerable interest due to its high accuracy. In this talk, we present our recent work on using deep learning for fingerprinting based indoor localization where Channel State Information (CSI), such as amplitude and phase difference information, are exploited for location estimation. Specifically, we present the design of ResLoc, which employs bi-modal CSI tensor data to train a deep residual sharing learning network. We then present DeepMap, a deep Gaussian process based approach for indoor radio map construction and location estimation, aiming to greatly reduce the training burden. Experimental results are presented to confirm that with deep learning and CSI, the proposed system can effectively reduce location error compared with existing methods in representative indoor environments.