Presenting the state of the art in several speech-related technologies, this book may be considered a follow-up to a previously published title [1]. The style is different, since this one is an edited volume and not tutorial in nature.

The book has 17 chapters, split into six parts. Following an introduction, Part 1 consists of two chapters devoted to speech quality assessment. The emphasis is on the impact on quality of speech bandwidth increase and methodologies for automated quality monitoring.

Part 2 is dedicated to adaptive algorithms for speech processing, and has four chapters. In the first one, a detailed analysis of a novel control methodology using Kalman filtering for high-performance acoustic echo cancellation is presented. The next chapter investigates state-of-the-art single-channel noise-reduction systems, of the spectral-subtraction type. Analysis of the well-known music noise artifacts in these systems is presented, as well as a method to control the audibility of these artifacts. The next chapter presents a comprehensive overview of acoustic source localization methods with microphone arrays, and also provides comparative simulation-based results. The last chapter of this part introduces sequences with perfect correlation properties, including their use in the identification of multiple-input multiple-output (MIMO) systems.

In Part 3, on speech coding, the emphasis is on technologies that enhance the perception of speech. This part has two chapters: the first is a detailed overview, and the second presents backwards-compatible wideband speech techniques.

The four chapters in Part 4 are devoted to essentially teaching how the merging of speech and channel coding may lead to more efficient and higher-quality systems. Parameter estimators, soft-decision source decoding, optimized channel coding, and iterative source channel decoding are discussed.

Part 5 presents advanced microphone signal processing for hearing aids. This part has two chapters. The first chapter, after a short overview of binaural microphone technologies and a wireless commercial hearing aid, presents the latest in binaural-beamforming methods. The next chapter also discusses microphone beamforming, but includes a new concept in this area: noise reduction based on the auditory profile of the listener.

The last part deals with speech processing for human-machine interfaces. It has two chapters, devoted to such topics as speech recognition in adverse acoustic conditions and speaker classification for voice-dialog systems. In the speech-recognition related chapter, the authors show how to adapt hidden Markov model-based recognizers, to cope with increased noise and reverberation levels. In the speaker classification chapter, the discussion is on methodologies for voice-services personalization, based on the classification of the speaker into a certain target group (for example, age or gender).

Written at a very advanced level, this book is recommended to anyone interested in the details of the latest developments in the area of digital-speech enhancement and processing.