Acoustic Scene Analysis, Complex Modulations, and a New Form of Filtering

Explore advanced signal processing techniques for acoustic scene analysis through this seminar lecture delivered by Les Atlas from the University of Washington at Johns Hopkins University's Center for Language & Speech Processing. Delve into the complexities of acoustic scene analysis, examining how complex modulations can be utilized to understand and interpret audio environments. Learn about innovative filtering approaches that represent a new paradigm in signal processing, moving beyond traditional methods to address the challenges of separating and analyzing multiple sound sources in complex acoustic environments. Discover how these techniques can be applied to improve speech recognition, audio source separation, and environmental sound analysis. Gain insights into the mathematical foundations underlying complex modulation analysis and understand how these concepts translate into practical applications for acoustic signal processing. The presentation covers both theoretical frameworks and practical implementations, providing a comprehensive overview of cutting-edge research in computational auditory scene analysis and its implications for future developments in speech and audio processing technologies.