Enregistré dans:
Détails bibliographiques
Auteurs principaux: Kando, Shunsuke, Miyao, Yusuke, Naradowsky, Jason, Takamichi, Shinnosuke
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:https://arxiv.org/abs/2407.10118
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866929420094144512
author Kando, Shunsuke
Miyao, Yusuke
Naradowsky, Jason
Takamichi, Shinnosuke
author_facet Kando, Shunsuke
Miyao, Yusuke
Naradowsky, Jason
Takamichi, Shinnosuke
contents Traditional spoken language processing involves cascading an automatic speech recognition (ASR) system into text processing models. In contrast, "textless" methods process speech representations without ASR systems, enabling the direct use of acoustic speech features. Although their effectiveness is shown in capturing acoustic features, it is unclear in capturing lexical knowledge. This paper proposes a textless method for dependency parsing, examining its effectiveness and limitations. Our proposed method predicts a dependency tree from a speech signal without transcribing, representing the tree as a labeled sequence. scading method outperforms the textless method in overall parsing accuracy, the latter excels in instances with important acoustic features. Our findings highlight the importance of fusing word-level representations and sentence-level prosody for enhanced parsing performance. The code and models are made publicly available: https://github.com/mynlp/SpeechParser.
format Preprint
id arxiv_https___arxiv_org_abs_2407_10118
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Textless Dependency Parsing by Labeled Sequence Prediction
Kando, Shunsuke
Miyao, Yusuke
Naradowsky, Jason
Takamichi, Shinnosuke
Computation and Language
Traditional spoken language processing involves cascading an automatic speech recognition (ASR) system into text processing models. In contrast, "textless" methods process speech representations without ASR systems, enabling the direct use of acoustic speech features. Although their effectiveness is shown in capturing acoustic features, it is unclear in capturing lexical knowledge. This paper proposes a textless method for dependency parsing, examining its effectiveness and limitations. Our proposed method predicts a dependency tree from a speech signal without transcribing, representing the tree as a labeled sequence. scading method outperforms the textless method in overall parsing accuracy, the latter excels in instances with important acoustic features. Our findings highlight the importance of fusing word-level representations and sentence-level prosody for enhanced parsing performance. The code and models are made publicly available: https://github.com/mynlp/SpeechParser.
title Textless Dependency Parsing by Labeled Sequence Prediction
topic Computation and Language
url https://arxiv.org/abs/2407.10118