Recent advances in speech recognition

Summary form only given. Speech recognition systems tend to consist of the following modules: input (acquisition and acoustic processing); classification (phoneme/word recognition); language (lexical/syntactic/semantic) processing; output (actions to be performed); although in many systems these modules are further subdivided, in some systems some of these modules may be omitted or several may be combined.