DETECTING AND RECOVERING OUT OF VOCABULARY WORDS IN SPEECH-TO-TEXT TRANSCRIPTION SYSTEMS

摘要

Aspects of the present disclosure describe methods for identifying and recovering out-of-vocabulary words in transcripts of a speech data recording using speech recognition models and phrase unit recognition models. An example method generally includes receiving a voice data recording for transcription into a textual representation of the voice data recording. The speech data record is transcribed into the text representation using a word recognition model. An unknown word is identified in the text representation and the unknown word is reconstructed based on a recognition of sub-units of the unknown word generated by a sub-unit recognition model. The textual representation of the speech data record is modified by replacing the unknown word with the reconstruction of the unknown word, and the modified textual representation is output.