BCN2BRNO Automatic speech recognition system for Albayzin 2022 Speech to Text Challenge

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F22%3APR37717" target="_blank" >RIV/00216305:26230/22:PR37717 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.fit.vut.cz/research/product/797/" target="_blank" >https://www.fit.vut.cz/research/product/797/</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

BCN2BRNO Automatic speech recognition system for Albayzin 2022 Speech to Text Challenge

Popis výsledku v původním jazyce

The software is based on the development of Automatic Speech Recognition systems for the Albayzin 2022 Challenge. We trained and evaluated both hybrid systems and those based on end-to-end models. We also investigated the use of self-supervised learning speech representations from pre-trained models and their impact on ASR performance (as opposed to training models directly from scratch). Additionally, we also applied the Whisper model in a zero-shot fashion, postprocessing its output to fit the required transcription format. On top of tuning the model architectures and overall training schemes, we improved the robustness of our models by augmenting the training data with noises extracted from the target domain. Moreover, we applied rescoring with an external LM on top of N-best hypotheses to adjust each sentence score and pick the single best hypothesis. All these efforts lead to a significant WER reduction. Our single best system and the fusion of selected systems achieved 16.3% and 13.7% WER respectively on RTVE2020 test partition, i.e. the official evaluation partition from the previous Albayzin challenge

Název v anglickém jazyce

BCN2BRNO Automatic speech recognition system for Albayzin 2022 Speech to Text Challenge

Popis výsledku anglicky

The software is based on the development of Automatic Speech Recognition systems for the Albayzin 2022 Challenge. We trained and evaluated both hybrid systems and those based on end-to-end models. We also investigated the use of self-supervised learning speech representations from pre-trained models and their impact on ASR performance (as opposed to training models directly from scratch). Additionally, we also applied the Whisper model in a zero-shot fashion, postprocessing its output to fit the required transcription format. On top of tuning the model architectures and overall training schemes, we improved the robustness of our models by augmenting the training data with noises extracted from the target domain. Moreover, we applied rescoring with an external LM on top of N-best hypotheses to adjust each sentence score and pick the single best hypothesis. All these efforts lead to a significant WER reduction. Our single best system and the fusion of selected systems achieved 16.3% and 13.7% WER respectively on RTVE2020 test partition, i.e. the official evaluation partition from the previous Albayzin challenge

Druh

R - Software

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/LTAIN19087" target="_blank" >LTAIN19087: Multi-lingualita v řečových technologiích</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Interní identifikační kód produktu

R1-LTAIN19087

Technické parametry

Pro stažení kontaktujte: Vysoké učení technické v Brně, IČ: 00216305, Ing. Martin Kocour, Fakulta Informační technologií, Božetěchova 2/1 612 00 Brno, tel.: 541141283, ikocour@fit.vut.cz, https://www.fit.vut.cz/person/ikocour/

Ekonomické parametry

Produkt vznikl v rámci vývojově-výzkumné činnosti na FIT VUT v Brně, zejména díky česko-indické spolupráci financované z projektu: LTAIN19087 - Multi-lingualita v řečových technologiích.

IČO vlastníka výsledku

00216305

Název vlastníka

Vysoké učení technické v Brně