Select a calendar:
Filter October Events by Event Type:
Events for the 3rd week of October
-
Fall 2018 Joint CSC@USC/CommNetS-MHI Seminar Series
Mon, Oct 15, 2018 @ 02:00 PM - 03:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Stanley Chan, Purdue University
Talk Title: Understanding Plug-and-Play ADMM: Convergence, Objective Function, and Generalization
Abstract: The Plug-and-Play (PnP) ADMM algorithm is a recently developed image restoration method that allows advanced image denoisers to be integrated into physical forward models to yield a provable convergent algorithm. Since its introduction in 2013, PnP ADMM has enabled numerous record-breaking image recovery results in deblurring, inpainting, super-resolution, Poisson denoising, and compressed sensing, etc. However, despite the successful applications and promising results, very little is known about why PnP ADMM performs so well. Fundamentally, the challenge lies in the fact that many of the latest denoisers are not easily expressible as proximal maps, e.g., deep neural networks.
In this talk, I will highlight a few recent progresses made by my group and collaborators. I will discuss three questions. (1) Convergence: Under what conditions of the denoisers will PnP ADMM converge? Answering this question will allow us to comment on what kind of denoisers we can use and what kind of convergence we should expect. (2) Objective Function: By plugging in an off-the-shelf denoiser, what does PnP ADMM actually solve? That is, what is the corresponding objective function? This problem will tell us why and when PnP ADMM will perform well, and when PnP ADMM will fail. (3) Generalization: Are we able to generalize PnP ADMM to accommodate multiple agents beyond a single forward model and a single denoiser? This leads to a new concept called consensus equilibrium, which allows us to integrate multiple weak experts to produce an overall strong recovery method. I will illustrate the ideas through examples in image denoising, graph signal processing, turbulence removal and automatic foreground extraction.
Biography: Stanley H. Chan is currently an Assistant Professor in the School of Electrical and Computer Engineering and the Department of Statistics at Purdue University, West Lafayette, IN. He received the B.Eng. degree in Electrical Engineering from the University of Hong Kong in 2007, the M.A. degree in Mathematics from UC San Diego in 2009, and the Ph.D. degree in Electrical Engineering from UC San Diego in 2011. In 2012-2014, he was a postdoctoral research fellow at Harvard. His PhD study and postdoctoral training were supported by the Croucher Foundation PhD Scholarship and postdoctoral Fellowship, two of the most prestigious scholarships in Hong Kong.
Dr. Chan is a recipient of the Best Paper Award of IEEE International Conference on Image Processing 2016 for his work on single photon image sensors. He is also a recipient of multiple education awards including the IEEE Signal Processing Cup Second Prize, Purdue College of Engineering Outstanding Graduate Mentor Award, Eta Kappa Nu Teaching Award, Eta Kappa Nu Outstanding Professor Award, and Purdue Teaching for Tomorrow Fellowship.
Dr. Chan is an Associate Editor of IEEE Transactions on Computational Imaging since 2018, an Associate Editor of OSA Optics Express in 2016 - 2018, an Elected Member and the subcommittee Chair of the IEEE Signal Processing Society Special Interest Group in Computational Imaging since 2015. He was the co-chair and co-organizer of the computational imaging special session in ICIP 2016, and had served on multiple technical program committees including ICIP, ICASSP, OSA Imaging and Applied Optics Congress, and Midwest Machine Learning Symposium.
Host: Antonio Ortega, antonio.ortega@sipi.usc.edu
More Info: http://csc.usc.edu/seminars/2018Fall/chan.html
More Information: 18.10.15 Stanley Chan_CSC@USC Seminar.pdf
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 132
Audiences: Everyone Is Invited
Contact: Brienne Moore
Event Link: http://csc.usc.edu/seminars/2018Fall/chan.html
-
Finding Structure in Data: Clustering and Representation Learning
Tue, Oct 16, 2018 @ 10:00 AM - 11:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Arya Mazumdar, College of Information & Computer Sciences, University of Massachusetts Amherst
Talk Title: Finding Structure in Data: Clustering and Representation Learning
Abstract: This talk is loosely divided into two parts, both about uncovering hidden structures in data by unsupervised or semisupervised methods. In the first, we discuss new tools to learn parameters of mixtures of distributions, statistical block models, and interactive algorithms for such problems. In the second, we describe new algorithms to learn nonlinear models of data, primarily focusing on networks of rectified linear units. We will emphasize on the information theoretic tools that have been used in both of the parts. We provide rigorous theoretical guarantees and our algorithms perform very well in experiments conducted with real data.
Biography: Arya Mazumdar is an assistant professor in the College of Information and Computer Sciences at the University of Massachusetts Amherst since Fall 2015. Prior to this, Arya was an assistant professor at the University of Minnesota-Twin Cities, and a postdoctoral scholar at Massachusetts Institute of Technology. Arya received his Ph.D. from the University of Maryland, College Park, in 2011, where his thesis won the Distinguished Dissertation Fellowship Award. Arya is a recipient of the 2015 NSF CAREER award and the 2010 IEEE ISIT Jack K. Wolf Student Paper Award. He spent the summers of 2008 and 2010 at the Hewlett-Packard Laboratories, Palo Alto, CA, and IBM Almaden Research Center, San Jose, CA, respectively. Arya's research interests include information theory and machine learning.
Host: Professor Salman Avestimehr
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Talyia White
-
Machine-Integrated Intelligence, Controlled Sensing, and Active Learning
Wed, Oct 17, 2018 @ 12:00 AM - 01:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Tara Javidi, Electrical and Computer Engineering, UC San Diego
Talk Title: Machine-Integrated Intelligence, Controlled Sensing, and Active Learning
Series: Center for Cyber-Physical Systems and Internet of Things
Abstract: The computing landscape has been drastically changing. The new computing realm, which is sometimes dubbed as internet of everything, includes networked devices ranging from tiny wearable sensors, smart home appliances, and personal autonomous robots, to connected self-driving cars, and to smart city infrastructures. In this new computing eco-system, comprising of resource-constrained, unreliable, and vulnerable components and networks, the non-recurring cost of hardware acceleration, engineering implementation, and system building has continued to grow significantly. This is in addition to the growing cost associated with the collection, curation, and labeling of data during both the training and the execution of various popular machine learning models. These design bottle-necks not only result in a significant increase in the non-recurring cost of engineering for companies but also provide a severe hurdle in technology development associated with hardware upgrade and/or system redesign.
In the first part of the talk, I will discuss an overview of my research on information acquisition and active learning in the context of the mission of our newly formed UCSD Center for Machine-Integrated Computing and Security (MICS). I will report ongoing research in the center where this system integrated view has enabled best-in-class results by bringing Machine into Machine Learning. In the second part of the talk, I will delve deeper into the problems of information acquisition, controlled sensing, and active learning and show our solutions to significantly reduce the cost of data collection and/or data labeling while ensuring reliability and fidelity during the training or run-time. In particular, we illustrate our findings and algorithms in the context of DetecDrone: an ML-enabled drone intelligence platform developed in my lab to provide search, mapping, and monitoring off-the-shelf low cost drones.
Biography: Tara Javidi studied electrical engineering at Sharif University of Technology, Tehran, Iran from 1992 to 1996. She received her MS degrees in electrical engineering (systems) and in applied mathematics (stochastic analysis) from the University of Michigan, Ann Arbor, in 1998 and 1999, respectively. She received her Ph.D. in electrical engineering and computer science from the University of Michigan, Ann Arbor, in 2002. From 2002 to 2004, Tara Javidi was an assistant professor at the Electrical Engineering Department, University of Washington, Seattle. In 2005, she joined the University of California, San Diego, where she is currently a professor of electrical and computer engineering and a founding co-director of the Center for Machine-aware Computing and Security (MICS). She is also a member of Board of Governors of the IEEE Information Theory Society (2017/18/19).
Tara Javidi's research interests are in theory of active learning, information theory with feedback, stochastic control theory, and stochastic resource allocation in wireless communications and communication networks. Tara Javidi was a recipient of a 2018 Qualcomm Faculty Award, National Science Foundation early career award (CAREER) in 2004, Barbour Graduate Scholarship, University of Michigan, in 1999, and the Presidential and Ministerial Recognitions for Excellence in the National Entrance Exam, Iran, in 1992. Tara Javidi is a Distinguished Lecturer of the IEEE Information Theory Society (2017/18).
Host: Professor Paul Bogdan
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 132
Audiences: Everyone Is Invited
Contact: Talyia Whtie
-
Stochastic Control of Finite and Infinite Dimensional Systems Under Uncertainty: Theory, Algorithms and Applications
Thu, Oct 18, 2018 @ 11:00 AM - 12:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Evangelos A. Theodorou , Guggenheim School of Aerospace Engineering, Georgia Institute of Technology
Talk Title: Stochastic Control of Finite and Infinite Dimensional Systems Under Uncertainty: Theory, Algorithms and Applications
Series: Center for Cyber-Physical Systems and Internet of Things
Abstract: In this talk, I will present an overview of projects related to stochastic control and machine learning methods and their applications to dynamical systems represented by stochastic differential and stochastic partial differential equations. These are typically systems that exists in autonomy and robotics as well as in areas of applied physics such as fluid mechanics, plasma physics and turbulence. I will discuss different forms of uncertainty representation that span Gaussian Processes, Polynomial Chaos, Deep Probabilistic Neural Networks and Q-Wiener processes. Finally, I will show applications to robotic terrestrial agility, perceptual control, social networks, large-scale swarms, and control of stochastic fields, and conclude with future directions.
Biography: Evangelos A. Theodorou is an assistant professor with the Guggenheim School of aerospace engineering at Georgia Institute of Technology. He is also affiliated with the Institute of Robotics and Intelligent Machines. Evangelos Theodorou earned his Diploma in Electronic and Computer Engineering from the Technical University of Crete (TUC), Greece in 2001. He has also received a MSc in Production Engineering from TUC in 2003, a MSc in Computer Science and Engineering from University of Minnesota in spring of 2007 and a MSc in Electrical Engineering on dynamics and controls from the University of Southern California(USC) in Spring 2010. In May of 2011 he graduated with his PhD, in Computer Science at USC. After his PhD, he was a Postdoctoral Research Fellow with the department of computer science and engineering, University of Washington, Seattle. Evangelos Theodorou is the recipient of the King-Sun Fu best paper award of the IEEE Transactions on Robotics for the year 2012 and recipient of the best paper award in cognitive robotics in International Conference of Robotics and Automation 2011. He was also the finalist for the best paper award in International Conference of Humanoid Robotics 2010, International Conference of Robotics and Automation 2017 and Robotics Science and Systems 2018. His theoretical research spans the areas of stochastic optimal control theory, machine learning, information theory and statistical physics. Applications involve learning, planning and control in autonomous, robotics and aerospace systems.
Host: Prof. Paul Bogdan
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Talyia White