SUNMONTUEWEDTHUFRISAT
Events for February 20, 2019
-
Center for Cyber-Physical Systems and Internet of Things and Ming Hsieh Institute Seminar Series
Wed, Feb 20, 2019 @ 03:00 PM - 04:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Farshad Lahouti, California Institute of Technology
Talk Title: Learning from Oracle or Crowd: Budget fidelity trade-offs and query design strategies
Series: Center for Cyber-Physical Systems and Internet of Things
Abstract: Design of many machine learning systems rely on well-curated datasets. Developing such datasets is an expensive and time-consuming process. Digital crowdsourcing (CS) is a modern approach to infer this from small contributions of a large and potentially non-expert crowd. In this talk, the CS problem, as a human-in-the-loop computation problem, is modeled and analyzed in an information theoretic rate-distortion framework. The purpose is to identify the ultimate fidelity that one can achieve by any form of query from the crowd and any inference algorithm with a given budget. This in turn motivates the design of coded query schemes. Strategies are presented for efficient and reliable query design in presence of a crowd or an oracle. The query rate performance and speed of learning are analyzed and the role of pricing is investigated. Joint work with Victoria Kostina and Babak Hassibi.
Biography: Farshad Lahouti received the Ph.D. degree in electrical engineering from the University of Waterloo, Canada, in 2002. In 2005, he joined the faculty of the School of Electrical and Computer Engineering, University of Tehran, where he founded the Center for Wireless Multimedia Communications. Dr Lahouti received the distinguished scientist award from Iran National Academy of Sciences in 2014. He joined the electrical engineering department at Caltech as a visiting faculty in 2013, where he initiated the digital ventures design program. He is also the co-founder and the scientific lead of cntxts Inc inventing the next generation ultra energy efficient IoT technology and its digital AI contexts. His current research interests are coding and information theory, and statistical signal processing with applications to machine learning, wireless networks and biological and neuronal networks. Information on his recent works can be found here: http://www.its.caltech.edu/~lahouti
Host: Salman Avestimehr
Location: Hughes Aircraft Electrical Engineering Center (EEB) - EEB 132
Audiences: Everyone Is Invited
Contact: Talyia White
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
AI, Biometrics and Computer Vision @ VISTA
Wed, Feb 20, 2019 @ 03:15 PM - 04:15 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Dr. Wael AbdAlmageed, USC-ISI
Talk Title: AI, Biometrics and Computer Vision @ VISTA
Abstract: In this talk, I will describe various ongoing research activities at computer vision group at USC Information Sciences Institute that span core machine learning and artificial intelligence, biometrics, image integrity assessment and fake news detection. In terms of core machine learning, I will present a novel representation (i.e. feature) learning framework for neural networks, called unsupervised adversarial invariance, which learns a split representation of data through competitive learning between a classification task and a reconstruction task coupled with information disentanglement, without needing any labeled information about confounding factors or domain knowledge. Second, in biometrics, I will present the design of a novel biometrics sensor suite that provides rich data for presentation attack detection (i.e. anti-spoofing) for face, iris and fingerprint biometrics. I will also present a deep neural network approach for presentation attack detection using limited amounts of data obtained from the new sensor suite. Finally, I will present two new approaches for assessing the integrity of multimedia assets. The first approach is called BusterNet, in which I will describe a novel convolutional neural network architecture for detection and localization of copy-move image forgeries. The second approach uses joint multimodal embedding to assess the integrity of multimedia packages undergoing semantic manipulations. I will conclude my talk with directions and plans for future research.
Biography: Dr. Wael AbdAlmageed is a Research Team Leader and Supervising Computer Scientist at the USC Information Sciences Institute (USC ISI). He is the Co-Director of the Center for Video, Image, Speech and Text Analytics (VISTA). He is an expert in computer vision, biometrics and machine learning. His research focus is on developing and applying artificial intelligence techniques to vision, bioinformatics and data analytics problems. His research interests also include mapping machine learning and computer vision algorithms to modern high performance computing platforms. Prior to joining ISI, from 2004 to 2013, Dr. AbdAlmageed was a research scientist with the University of Maryland at College Park, where he led research and development efforts for various DARPA, IARPA and ARL programs, such as VIVID, VIRAT, PerSeas and VACE. He obtained his Ph.D. with Distinction from the University of New Mexico in 2003 where he was also awarded the Outstanding Graduate Student award. He has two patents and over 65 publications in top machine learning, computer vision and high performance computing conferences and journals. Dr. AbdAlmageed is the PI for ISI's effort on IARPA Odin, and DARPA MediFor.
Host: Profs Richard Leahy & Shrikanth Narayanan
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Mayumi Thrasher
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor.
