This is another in the Departmental Pizza Seminar series.

This is a CSMS Project Industrial Liaison Seminar. Refreshments will be served.

Over the last forty years, remarkable progress has been made in the area of speech separation and enhancement, however accurate estimation of clean speech in real-world environments is still a challenge. We address the problem of speech estimation as statistical estimation with "missing" data in the independent component analysis (ICA) domain. Missing components are substituted by values drawn from "similar" data in a multi-faceted ICA representation of the complete data. I will present the algorithm for the inference of missing data in the case of a fixed pattern of missing data, and then I will present an application of our approach to the problem of bandwidth extension, where speech is degraded by a fixed filtering process. I will show the capability of the algorithm to reconstruct fine missing details of the original data with little artifacts: information of the source signal can indeed be modeled and used in order to recreate a natural sounding source in adverse conditions. The extension of the method to statistical spectral inference according to random patterns of missingness promises progress in the long open problem of performing speech enhancement while enhancing the intelligibility of speech.

This work has been done in collaboration with Doru-Cristian Balcan (Carnegie Mellon University) and Timo Gerkmann (Bochum University).

Biography: Justinian Rosca is Program Manager in Audio, Signal Processing and Wireless Communications at Siemens Corporate Research in Princeton, USA. He is also Affiliate Professor, Department of Electrical Engineering of University of Washington, Seattle, USA. He received the Dipl. Eng. degree in Computers and Control Engineering from Bucharest Polytechnic University in 1984, the M.S. and Ph.D. degrees in Computer Science from University of Rochester in 1992 and 1997 respectively. Dr. Rosca is conducting research in statistical signal processing and radio management, with an emphasis on topics involving acquisition, management and processing of data with uncertainties, such as audio processing, blind signal separation, wireless management, adaptive principles in stochastic search and optimization, and probabilistic inference in artificial intelligence.

For new terminals such as Network PC and thin-client, broadband network access is becoming significant. For mobile users, wireless communication technology such as Cognitive Radio (CR) is expected to realize seamless communications. However, QoS consideration in CR has not been well investigated. This talk intorduces a QoS control framework which consists of such as optimization of resource allocation, and shows a result of "Geographical Optimal Route Selection with Cost-Constraint".

This is another in the Departmental Pizza Seminar series.

This is another in the Departmental Pizza Seminar series.

An ontology is a formal, explicit specification of a conceptualization. The process of engineering an ontology for a domain using the top-down approach is a complex task that consumes a lot of time and effort (it's the knowledge acquisition bottleneck). When the domain is a substantially large amount of texts (corpus), one way to expedite this prior process is to reverse engineer the ontology from the corpus. In this talk I am presenting my master's research proposal on an efficient and effective method to solve the reverse engineering problem using lexico-semantic analysis and probabilistic reasoning.

This is another in the Departmental Pizza Seminar series.

Almost everyone has to give a public presentation at some time. Graduate students have to defend their theses, research students have to present their results, and many jobs require presentations. It is common to have little or no experience when you give your first presentation, and you may even be a little nervous! This talk is aimed at (graduate, research, and other) students. It describes how to present a successful talk, in a simple standard format. This talk does not try to teach general speaking skills, nor impose any personally preferred techniques. It covers the structure of a talk, the use of visual aids, speaking technique, and how to cope with questions.

This is another in the Departmental Pizza Seminar series.

In the first part of this talk we defined the problem we want to solve, i.e. the need of a systematic and flexible approach to perform pattern classification of biological data coming from many sources. Since the nature of the problem dictates the nature of the solution, in this talk we will talk about the scope of the BCD, the limitations found during its implementation, and how they were resolved. The present state of the BCD and its future direction will be discussed. The bottom question is: how general can an information system be? As we will see, not much, i.e. 'very general' with a twist.

This is another in the Departmental Pizza Seminar series.

Subduction zones, where oceanic plates are pushed under the leading edge of continental plates along ocean trenches, produce Earth's largest earthquakes and most tsunamis. The pattern of strain release during earthquakes, where the leading edge of the continental plate jumps towards the ocean by several meters or more, is related to the slow build-up of strain accumulation during the interseismic period, which may last for hundreds of years. New GPS technology permits this slow pattern of strain accumulation to be measured. Studies of strain accumulation may give clues to the nature of future earthquakes, leading to improved understanding of the seismic process and improved forecast of seismic hazard. However, GPS data at a number of subuction zones indicates that not all accumulated strain is released during earthquakes; slow, aseismic slip events with durations of days - months are increasingly recognized as a major component in the strain release budget. In this talk I will describe a new GPS and seismic network that is being installed in northern Costa Rica by the University of Miami to monitor such events, and describe preliminary results from the first three years of operation. We have already observed one slow slip event, in May 2007. Maximum surface offsets were approximately 2 cm, occurring over a duration of several weeks, corresponding to an ~ M 6.5 earthquake if all of this strain had been released rapidly. Maximum slip was centered near the down-dip edge of the conventionally defined seismogenic zone. How these data are collected, analyzed and interpreted will be discussed, as well as implications for future earthquakes and improved understanding of the earthquake process.

This is a CSMS Project Industrial Liaison Seminar. Refreshments will be served.

Biodatabases are becoming increasingly accessible, increasingly big, and increasingly heterogeneous. The BCD is an (ongoing) open web information system that provides a systematic and flexible approach to perform pattern classification of biological data coming from many sources. In this talk I will describe the practical problems encountered during the design and development of the BCD and the solutions implemented related to high performance, high availability, high maintainability, high extensibility, cloud computing, database federation, web services, configuration management, and integration of third-party calculated feature vectors into the BCD data mining framework. The present state of the BCD and its future direction will be discussed. The bottom question is: how general can an information system be? As we will see, not much, i.e. "very general" with a twist.

This is another in the Departmental Pizza Seminar series.

In this talk, I will discuss the role of mathematics and computer science at NSA and discuss hiring opportunities, including REUs (Research Experiences for Undergraduates) and programs for graduate students.

This is a CSMS Project Industrial Liaison Seminar. Refreshments will be served.

Autonomous agents in real-time and dynamic adversarial environments offer numerous research challenges. Perception, localization, decision- making, communication, and locomotion are good examples. The novel modern sports simulator we will discuss integrates results from ten years of research in the area of autonomous soccer playing robots (both softbots and physical robots) with RoboCup as a testbed. We will explore the problem of enabling autonomous agents in finding the right passing point or in making a complex decision within a soccer game while dealing with time constrains, hostile opponents, and dynamic environments. We propose a framework for spatio-temporal real-time analysis of dynamic scenes. The underlying hierarchical three-tier multiagent system consists of autonomous BDI agents that allows dynamic group structures (e.g., an emergent situation for a wing attack). The online game runs seamlessly in a web browser with a new and state-of-the-art 3D visualization engine. Latest developments include research results from a motion capturing lab and face generators to enhance the believability of the players and the users' visualization experience.

This is another in the Departmental Pizza Seminar series.

Web Services is one of the most dominant tools in the world today to implement SaaS (Software As A Service) principles. It has revolutionized the way services are mashed-up to build complex services and governing bodies. QoS (Quality of Services) such as security, reliability, etc., plays an important role in creating an ultimate service with governance. This presentation will focus on harnessing the power of Web Services using Apache Axis2 to create such a service and invoking prior service using Apache Axis2 client in a matter of minutes. In logic ...

This is another in the Departmental Pizza Seminar series.

There is an enormous demand from the industry for Web Services and related technologies at present. These demands are not only from the IT field, but also from other disciplines such as finance, telecommunications, and government. Web Services is one of the most successful implementations of Service Oriented Architectures, which provides an infrastructure to build reusable software services in a heterogeneous environment. Apache Axis2 is considered to be the most successful open source Web Services middleware platform offered by the Apache Software Foundation. This presentation will focus on the basics of the Apache Axis2 architecture, concepts, and R&D opportunities available for students in the open source arena.

This is another in the Departmental Pizza Seminar series.

In recent years there has been a growing demand for Automated Theorem Proving (ATP) in large theories, which often have more axioms than can be handled effectively as normal internal axioms. This work addresses the issues of accessing external sources of axioms from a first-order logic ATP system, and presents an implemented ATP system that retrieves external axioms asynchronously, on demand. The efficacy of the system is demonstrated on test problems that make use of a range of external sources of axioms, including databases and web services. In the long term this system will be able to answer very general knowledge questions, using a natural language interface.

This is another in the Departmental Pizza Seminar series.