The performance of conventional Automatic Speech Recognition(ASR) systems degrades dramatically as soon as the microphone ismoved away from the mouth of the speaker. This is due to a broadvariety of effects such as background noise, overlapping speechfrom other speakers, and reverberation. While traditional ASRsystems underperform for speech captured with far-field sensors,there are a number of novel techniques within the recognitionsystem as well as techniques developed in other areas of signalprocessing that can mitigate the deleterious effects of noise andreverberation, as well as separating speech from overlappingspeakers.

Distant Speech Recognitionpresents a contemporary andcomprehensive description of both theoretic abstraction andpractical issues inherent in the distant ASR problem.

Key Features:

• Covers the entire topic of distant ASR and offers practicalsolutions to overcome the problems related to it
• Provides documentation and sample scripts to enable readers toconstruct state-of-the-art distant speech recognition systems
• Gives relevant background information in acoustics and filtertechniques,
• Explains the extraction and enhancement of classificationrelevant speech features
• Describes maximum likelihood as well as discriminativeparameter estimation, and maximum likelihood normalizationtechniques
• Discusses the use of multi-microphone configurations forspeaker tracking and channel combination
• Presents several applications of the methods and technologiesdescribed in this book
• Accompanying website with open source software and tools toconstruct state-of-the-art distant speech recognition systems

This reference will be an invaluable resource for researchers,developers, engineers and other professionals, as well as advancedstudents in speech technology, signal processing, acoustics,statistics and artificial intelligence fields.