Uncertainty estimation for a speech recognition system

 

Αποθηκεύτηκε σε:
Λεπτομέρειες βιβλιογραφικής εγγραφής
Συγγραφείς: Morales-Muñoz, Walter, Calderón-Ramírez, Saúl
Μορφή: artículo original
Κατάσταση:Versión publicada
Ημερομηνία έκδοσης:2024
Περιγραφή:Whisper is a voice recognition system designed by the company OpenAI, which has been trained with 680,000 hours of multilingual and multitask supervised data collected from the web. The following research aims to adapt and employ the Monte Carlo Dropout using audio data labeled in Spanish and contaminated with a certain amount of noise and Levensthein distance to estimate the score uncertainty of this system.Preliminary results show that there is a linear relationship between uncertainty estimation and the Word Error Rate (WER) of the transcriptions. Furthermore, it is observed that the number of insertions or omissions in the transcriptions tends to be low.
Χώρα:Portal de Revistas TEC
Ίδρυμα:Instituto Tecnológico de Costa Rica
Repositorio:Portal de Revistas TEC
Γλώσσα:Español
OAI Identifier:oai:ojs.pkp.sfu.ca:article/7305
Διαθέσιμο Online:https://revistas.tec.ac.cr/index.php/tec_marcha/article/view/7305
Λέξη-Κλειδί :Uncertainty
Speech Recognition
ASR
Whisper
Monte Carlo Dropout
Incertidumbre
Reconocimiento de voz