It is generally agreed that speech will play a major role in defining next-generation human-machine interfaces because it is the most natural means of communication among humans. To push forward this vision, speech research has enjoyed a long and glorious history spanning the entire twentieth century. As a result in the last three decades we have witnessed an intensive technology progress spurred on by recent advances in speech modeling, coordinated efforts between government funding agencies and speech communities for data collection and benchmark performance evaluation, and easy accesses to fast and affordable computing machineries. In the context of spoken language processing, we consider a collection of technical topics ranging over all aspects of speech communication, including production, perception, recognition, verification, synthesis, coding, analysis, and modeling. We have also seen quite a few spoken language system concepts moving out of research laboratories, and being deployed into real-life services and applications.
