Language-Universal Speech Attributes Modeling for Zero-Shot Multilingual Spoken Keyword Recognition

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yen, Hao, Ku, Pin-Jui, Siniscalchi, Sabato Marco, Lee, Chin-Hui
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913377082671104
author Yen, Hao
Ku, Pin-Jui
Siniscalchi, Sabato Marco
Lee, Chin-Hui
author_facet Yen, Hao
Ku, Pin-Jui
Siniscalchi, Sabato Marco
Lee, Chin-Hui
contents We propose a novel language-universal approach to end-to-end automatic spoken keyword recognition (SKR) leveraging upon (i) a self-supervised pre-trained model, and (ii) a set of universal speech attributes (manner and place of articulation). Specifically, Wav2Vec2.0 is used to generate robust speech representations, followed by a linear output layer to produce attribute sequences. A non-trainable pronunciation model then maps sequences of attributes into spoken keywords in a multilingual setting. Experiments on the Multilingual Spoken Words Corpus show comparable performances to character- and phoneme-based SKR in seen languages. The inclusion of domain adversarial training (DAT) improves the proposed framework, outperforming both character- and phoneme-based SKR approaches with 13.73% and 17.22% relative word error rate (WER) reduction in seen languages, and achieves 32.14% and 19.92% WER reduction for unseen languages in zero-shot settings.
format Preprint
id arxiv_https___arxiv_org_abs_2406_02488
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Language-Universal Speech Attributes Modeling for Zero-Shot Multilingual Spoken Keyword Recognition
Yen, Hao
Ku, Pin-Jui
Siniscalchi, Sabato Marco
Lee, Chin-Hui
Audio and Speech Processing
Computation and Language
Sound
We propose a novel language-universal approach to end-to-end automatic spoken keyword recognition (SKR) leveraging upon (i) a self-supervised pre-trained model, and (ii) a set of universal speech attributes (manner and place of articulation). Specifically, Wav2Vec2.0 is used to generate robust speech representations, followed by a linear output layer to produce attribute sequences. A non-trainable pronunciation model then maps sequences of attributes into spoken keywords in a multilingual setting. Experiments on the Multilingual Spoken Words Corpus show comparable performances to character- and phoneme-based SKR in seen languages. The inclusion of domain adversarial training (DAT) improves the proposed framework, outperforming both character- and phoneme-based SKR approaches with 13.73% and 17.22% relative word error rate (WER) reduction in seen languages, and achieves 32.14% and 19.92% WER reduction for unseen languages in zero-shot settings.
title Language-Universal Speech Attributes Modeling for Zero-Shot Multilingual Spoken Keyword Recognition
topic Audio and Speech Processing
Computation and Language
Sound
url https://arxiv.org/abs/2406.02488