SUNMONTUEWEDTHUFRISAT
Events for March 01, 2017
-
Detecting paralinguistic information from speech and language for clinical applications: Algorithms and information limits
Wed, Mar 01, 2017 @ 10:00 AM - 12:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Visar Berisha, Arizona State University
Talk Title: Detecting paralinguistic information from speech and language for clinical applications: Algorithms and information limits
Abstract: The ability to share our thoughts and ideas through spoken
communication is fragile. Even the simplest verbal response requires a
complex sequence of events. It requires thinking of the words that best
convey your message; sequencing these words appropriately; and then
sending signals to the muscles required to produce speech. The slightest
damage to the brain areas that orchestrate these events can manifest in
speech and language problems. These disturbances offer a window into
brain functioning. In the first part of this presentation, I will
present an overview of a number of projects where we use interpretable
measures of speech and language production as proxies for cognitive and
motor health. The algorithms behind this work have practical utility in
clinical applications and can help answer basic research questions
related to dysarthric speech production.
In the second part of the talk, I will discuss new results from
non-parametric statistical signal processing that allow us to
characterize the information limits in speech. In contrast to existing
methods based on machine learning, this work provides a framework to
answer fundamental questions such as 'What are the bounds on how well I
can recover a parameter of interest from speech?' or 'How well should an
optimally trained classifier work for a particular application?'
Biography: Visar Berisha is an Assistant Professor at Arizona State
University with a joint appointment in the School of Electrical Computer
and Energy Engineering and the Department of Speech and Hearing Science.
Prior to joining ASU, Berisha was a research scientist at MIT Lincoln
Laboratory and then Principal Research Engineer for a Fortune 500
company. His research interests include speech analytics, statistical
signal processing, and information theory. Much of his recent work spans
all three of these fields to answer basic questions related to the
limits of information in speech. His research has led to many academic
publications, several licensed patents, and a revenue-positive startup
company. Berisha's work has been featured in the Science section of the
New York Times, on National Public Radio, and a number of other national
media outlets.
Host: Shrikanth Narayanan
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 132
Audiences: Everyone Is Invited
Contact: Tanya Acevedo-Lam/EE-Systems
-
MHI CommNetS Seminar
Wed, Mar 01, 2017 @ 02:00 PM - 03:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Dr. Arpan Chattopadhyay, USC
Talk Title: Sequential decision algorithms for as-you-go deployment of wireless relay network along a line
Series: CommNetS
Abstract: We are motivated by the need, in some applications, for impromptu or as-you-go deployment of wireless sensor networks. A person walks along a line, starting from a sink node (e.g., a base-station), and proceeds towards a source node (e.g., a sensor) which is at an a priori unknown location. At equally spaced locations, he makes link quality measurements to the previous relay, and deploys relays at some of these locations, with the aim to connect the source to the sink by a multihop wireless path. In this paper, we consider two approaches for impromptu deployment: (i) the deployment agent can only move forward (which we call a pure as-you-go approach), and (ii) the deployment agent can make measurements over several consecutive steps before selecting a placement location among them (the explore-forward approach). We consider a very light traffic regime, and formulate the problem as a Markov decision process, where the trade-off is among the power used by the nodes, the outage probabilities in the links, and the number of relays placed per unit distance. We obtain the structures of the optimal policies for the pure as-you-go approach as well as for the explore-forward approach. We also consider natural heuristic algorithms, for comparison. Numerical examples show that the explore-forward approach significantly outperforms the pure as- you-go approach in terms of network cost. Next, we propose learning algorithms for the explore-forward approach and the pure as-you-go approach, based on single and two timescale Stochastic Approximation, which asymptotically converge to the set of optimal policies, without using any knowledge of the radio propagation model. We demonstrate numerically that the learning algorithms can converge (as deployment progresses) to the set of optimal policies reasonably fast and, hence, can be practical model-free algorithms for deployment over large regions. Finally, we demonstrate the end-to-end traffic carrying capability of such networks via field deployment.
Biography: Arpan Chattopadhyay obtained his B.E. in Electronics and Telecommunication Engineering from Jadavpur University, Kolkata, India in the year 2008, and M.E. and Ph.D in Telecommunication Engineering from Indian Institute of Science, Bangalore in the year 2010 and 2015, respectively. Then he worked as a postdoc in the group DYOGENE of INRIA/ENS Paris. He joined EE department, USC as a postdoc from November 2016. His host is Prof. Urbashi Mitra. His research interests include optimization, learning and control of wireless networks and cyber-physical systems.
Host: Prof. Ashutosh Nayyar
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Annie Yu
