Automatic speech recognition (ASR) refers to the task of extracting a transcription of the linguistic content of an acoustical speech signal automatically. Despite several decades of research in this important area of ...