German Chapter
Signal Processing
Signal Processing
"The Bootstrap Paradigm in Signal Processing: Estimation, Detection and Model Selection"
The use of more accurate models in signal processing applications such as communications, radar, sonar, biomedical engineering, speech and image processing and machine learning has become a fundamental requirement. With an improved accuracy the models have become more complex and inferential statistical signal processing required in parameter estimation and signal detection and classification, for example, has become intractable. The signal processing practitioner requires a simple but accurate method for assessing errors of estimates and answering inferential questions. Asymptotic approximations are useful only when enough data is available, which is not always possible due to time constraints, the nature of the signal or the measurement setting. In place of the formulae and tables of parametric and non-parametric procedures based on complicated mathematics and asymptotic approximations, tools such as the Bootstrap have revolutionized statistics in the last decade and have become powerful for solving complex engineering problems. It is the method of an engineer's choice for solving inferential signal processing problems, such as signal detection, confidence limits estimation and model selection, to mention a few.
First, a brief history of the bootstrap methodology is given. We then discuss the basic principle underlying the bootstrap method for identically and independently distributed data and give several practical examples of its use in estimation, signal detection and model selection. A comprehensive overview of the bootstrap for dependent data is also given with emphasis on spectral analysis. Examples with real-life measurements are given throughout the talk.
Time: April 1, 2010, 13.00 c.t.
Location: Helmut Schmidt University, University of the Federal Armed Forces, Hamburg, Building H1, 2nd floor, room 2430
Time: September 25, 2008, 17:00
Location: Helmut Schmidt University, Hamburg, Building H1, Room 2430
This seminar introduces a new smoother solution which is motivated by Wiener filtering and minimises the variance of the estimation error. The optimum minimum-variance solution involves a cascade of a Kalman predictor and an adjoint Kalman predictor. Time-varying continuous-time and discrete-time smoothers are described. A speech enhancement example is presented in which the minimum-variance and the standard maximum likelihood smoother exhibit the same performance.
Garry A. Einicke is an adjunct associate professor at the University of Queensland and a principal research scientist in CSIRO Exploration and Mining, where he is the leader of the signal processing and navigation team. Garry is a reviewer for the IEEE control, signal processing and communication Societies. He chairs the signal processing and communications chapter in the Queensland section of the IEEE, and can be contacted at CSIRO, Technology Court, Pullenvale 4069, Australia.
Time: July 24, 2008
Location: Technical University Darmstadt, S306/51
Time: April 11, 2008
Location: Siemens Audiologische Technik GmbH, Erlangen
Time: October 12, 2007, 8:30am - 4:30pm
Location: Fachhochschule Dortmund
The program included a talk by IEEE Distinguished Lecturer Walter Kellermann.
Time: June 25, 2007, 11:00 am - 3:45 pm
Location: Leibniz Universität Hannover, Institut für Theoretische Elektrotechnik, Appelstr. 9a, 16. Stock
