Like Humans Do - Speech Recognition Scene Analysis Book

Improve your understanding of automatic speech recognition with Like Humans Do. This technical book explores a systematic approach to recognizing simultaneous speech signals through computational auditory scene analysis. By drawing inspiration from human auditory perception, the text investigates specific algorithms and techniques designed to decompose multiple speech signals effectively. The content focuses on integrating a spectro-temporal fragment decoder within a statistical search process. This method provides a deep look into the mechanisms needed to help automatic speech recognition reach human levels of performance. It is a valuable resource for those studying computer science, technology, and applications who want to understand the transition from current systems to more robust, human-like auditory processing.