An approach to efficient generation of high-accuracy and compact error-corrective models for speech recognition.

Takanobu Oba, Takaaki Hori, Atsushi Nakamura

Browse the full Interspeech paper archive.