A Temporal Extension of Latent Dirichlet Allocation for Unsupervised Acoustic Unit Discovery

Latent Dirichlet allocation (LDA) is widely used for unsupervised topic modelling on sets of documents. No temporal information is used in the model. However, there is often a relationship between the corresponding topics of consecutive tokens. In this paper, we present an extension to LDA that uses...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: van der Merwe, Werner, Kamper, Herman, Preez, Johan du
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator van der Merwe, Werner
Kamper, Herman
Preez, Johan du
description Latent Dirichlet allocation (LDA) is widely used for unsupervised topic modelling on sets of documents. No temporal information is used in the model. However, there is often a relationship between the corresponding topics of consecutive tokens. In this paper, we present an extension to LDA that uses a Markov chain to model temporal information. We use this new model for acoustic unit discovery from speech. As input tokens, the model takes a discretised encoding of speech from a vector quantised (VQ) neural network with 512 codes. The goal is then to map these 512 VQ codes to 50 phone-like units (topics) in order to more closely resemble true phones. In contrast to the base LDA, which only considers how VQ codes co-occur within utterances (documents), the Markov chain LDA additionally captures how consecutive codes follow one another. This extension leads to an increase in cluster quality and phone segmentation results compared to the base LDA. Compared to a recent vector quantised neural network approach that also learns 50 units, the extended LDA model performs better in phone segmentation but worse in mutual information.
doi_str_mv 10.48550/arxiv.2206.11706
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2206_11706</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2206_11706</sourcerecordid><originalsourceid>FETCH-LOGICAL-a676-832a52f156db965657d80fc8a3c13ca34d3c3a9c821d47ae3651a2a8f1d268e73</originalsourceid><addsrcrecordid>eNotj8lqwzAYhHXpoaR9gJ6qF7CrxVpyNGm6gKEX91Ywf7VQgWIZyTHJ2zdOexpmhhn4EHqgpG60EOQJ8iksNWNE1pQqIm_RV4t7d5hShoj3p9mNJaQRJ487uJgZP4cczE90M25jTAbmtfYp48-xHCeXl1Ccxa1JxzIHc0nDuikmLS6f79CNh1jc_b9uUP-y73dvVffx-r5ruwqkkpXmDATzVEj7vZVCCmU18UYDN5Qb4I3lhsPWaEZto8BxKSgw0J5aJrVTfIMe_26veMOUwwHyeVgxhysm_wV6H04L</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>A Temporal Extension of Latent Dirichlet Allocation for Unsupervised Acoustic Unit Discovery</title><source>arXiv.org</source><creator>van der Merwe, Werner ; Kamper, Herman ; Preez, Johan du</creator><creatorcontrib>van der Merwe, Werner ; Kamper, Herman ; Preez, Johan du</creatorcontrib><description>Latent Dirichlet allocation (LDA) is widely used for unsupervised topic modelling on sets of documents. No temporal information is used in the model. However, there is often a relationship between the corresponding topics of consecutive tokens. In this paper, we present an extension to LDA that uses a Markov chain to model temporal information. We use this new model for acoustic unit discovery from speech. As input tokens, the model takes a discretised encoding of speech from a vector quantised (VQ) neural network with 512 codes. The goal is then to map these 512 VQ codes to 50 phone-like units (topics) in order to more closely resemble true phones. In contrast to the base LDA, which only considers how VQ codes co-occur within utterances (documents), the Markov chain LDA additionally captures how consecutive codes follow one another. This extension leads to an increase in cluster quality and phone segmentation results compared to the base LDA. Compared to a recent vector quantised neural network approach that also learns 50 units, the extended LDA model performs better in phone segmentation but worse in mutual information.</description><identifier>DOI: 10.48550/arxiv.2206.11706</identifier><language>eng</language><subject>Computer Science - Computation and Language ; Computer Science - Learning ; Statistics - Machine Learning</subject><creationdate>2022-06</creationdate><rights>http://creativecommons.org/licenses/by-sa/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2206.11706$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2206.11706$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>van der Merwe, Werner</creatorcontrib><creatorcontrib>Kamper, Herman</creatorcontrib><creatorcontrib>Preez, Johan du</creatorcontrib><title>A Temporal Extension of Latent Dirichlet Allocation for Unsupervised Acoustic Unit Discovery</title><description>Latent Dirichlet allocation (LDA) is widely used for unsupervised topic modelling on sets of documents. No temporal information is used in the model. However, there is often a relationship between the corresponding topics of consecutive tokens. In this paper, we present an extension to LDA that uses a Markov chain to model temporal information. We use this new model for acoustic unit discovery from speech. As input tokens, the model takes a discretised encoding of speech from a vector quantised (VQ) neural network with 512 codes. The goal is then to map these 512 VQ codes to 50 phone-like units (topics) in order to more closely resemble true phones. In contrast to the base LDA, which only considers how VQ codes co-occur within utterances (documents), the Markov chain LDA additionally captures how consecutive codes follow one another. This extension leads to an increase in cluster quality and phone segmentation results compared to the base LDA. Compared to a recent vector quantised neural network approach that also learns 50 units, the extended LDA model performs better in phone segmentation but worse in mutual information.</description><subject>Computer Science - Computation and Language</subject><subject>Computer Science - Learning</subject><subject>Statistics - Machine Learning</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotj8lqwzAYhHXpoaR9gJ6qF7CrxVpyNGm6gKEX91Ywf7VQgWIZyTHJ2zdOexpmhhn4EHqgpG60EOQJ8iksNWNE1pQqIm_RV4t7d5hShoj3p9mNJaQRJ487uJgZP4cczE90M25jTAbmtfYp48-xHCeXl1Ccxa1JxzIHc0nDuikmLS6f79CNh1jc_b9uUP-y73dvVffx-r5ruwqkkpXmDATzVEj7vZVCCmU18UYDN5Qb4I3lhsPWaEZto8BxKSgw0J5aJrVTfIMe_26veMOUwwHyeVgxhysm_wV6H04L</recordid><startdate>20220623</startdate><enddate>20220623</enddate><creator>van der Merwe, Werner</creator><creator>Kamper, Herman</creator><creator>Preez, Johan du</creator><scope>AKY</scope><scope>EPD</scope><scope>GOX</scope></search><sort><creationdate>20220623</creationdate><title>A Temporal Extension of Latent Dirichlet Allocation for Unsupervised Acoustic Unit Discovery</title><author>van der Merwe, Werner ; Kamper, Herman ; Preez, Johan du</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a676-832a52f156db965657d80fc8a3c13ca34d3c3a9c821d47ae3651a2a8f1d268e73</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Computer Science - Computation and Language</topic><topic>Computer Science - Learning</topic><topic>Statistics - Machine Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>van der Merwe, Werner</creatorcontrib><creatorcontrib>Kamper, Herman</creatorcontrib><creatorcontrib>Preez, Johan du</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv Statistics</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>van der Merwe, Werner</au><au>Kamper, Herman</au><au>Preez, Johan du</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>A Temporal Extension of Latent Dirichlet Allocation for Unsupervised Acoustic Unit Discovery</atitle><date>2022-06-23</date><risdate>2022</risdate><abstract>Latent Dirichlet allocation (LDA) is widely used for unsupervised topic modelling on sets of documents. No temporal information is used in the model. However, there is often a relationship between the corresponding topics of consecutive tokens. In this paper, we present an extension to LDA that uses a Markov chain to model temporal information. We use this new model for acoustic unit discovery from speech. As input tokens, the model takes a discretised encoding of speech from a vector quantised (VQ) neural network with 512 codes. The goal is then to map these 512 VQ codes to 50 phone-like units (topics) in order to more closely resemble true phones. In contrast to the base LDA, which only considers how VQ codes co-occur within utterances (documents), the Markov chain LDA additionally captures how consecutive codes follow one another. This extension leads to an increase in cluster quality and phone segmentation results compared to the base LDA. Compared to a recent vector quantised neural network approach that also learns 50 units, the extended LDA model performs better in phone segmentation but worse in mutual information.</abstract><doi>10.48550/arxiv.2206.11706</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2206.11706
ispartof
issn
language eng
recordid cdi_arxiv_primary_2206_11706
source arXiv.org
subjects Computer Science - Computation and Language
Computer Science - Learning
Statistics - Machine Learning
title A Temporal Extension of Latent Dirichlet Allocation for Unsupervised Acoustic Unit Discovery
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-24T19%3A31%3A16IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=A%20Temporal%20Extension%20of%20Latent%20Dirichlet%20Allocation%20for%20Unsupervised%20Acoustic%20Unit%20Discovery&rft.au=van%20der%20Merwe,%20Werner&rft.date=2022-06-23&rft_id=info:doi/10.48550/arxiv.2206.11706&rft_dat=%3Carxiv_GOX%3E2206_11706%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true