TalTech-IRIT-LIS Speaker and Language Diarization Systems for DISPLACE 2024

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kalda, Joonas, Alumäe, Tanel, Lebourdais, Martin, Bredin, Hervé, Baroudi, Séverin, Marxer, Ricard
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913434528907264
author Kalda, Joonas
Alumäe, Tanel
Lebourdais, Martin
Bredin, Hervé
Baroudi, Séverin
Marxer, Ricard
author_facet Kalda, Joonas
Alumäe, Tanel
Lebourdais, Martin
Bredin, Hervé
Baroudi, Séverin
Marxer, Ricard
contents This paper describes the submissions of team TalTech-IRIT-LIS to the DISPLACE 2024 challenge. Our team participated in the speaker diarization and language diarization tracks of the challenge. In the speaker diarization track, our best submission was an ensemble of systems based on the pyannote.audio speaker diarization pipeline utilizing powerset training and our recently proposed PixIT method that performs joint diarization and speech separation. We improve upon PixIT by using the separation outputs for speaker embedding extraction. Our ensemble achieved a diarization error rate of 27.1% on the evaluation dataset. In the language diarization track, we fine-tuned a pre-trained Wav2Vec2-BERT language embedding model on in-domain data, and clustered short segments using AHC and VBx, based on similarity scores from LDA/PLDA. This led to a language diarization error rate of 27.6% on the evaluation data. Both results were ranked first in their respective challenge tracks.
format Preprint
id arxiv_https___arxiv_org_abs_2407_12743
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle TalTech-IRIT-LIS Speaker and Language Diarization Systems for DISPLACE 2024
Kalda, Joonas
Alumäe, Tanel
Lebourdais, Martin
Bredin, Hervé
Baroudi, Séverin
Marxer, Ricard
Audio and Speech Processing
This paper describes the submissions of team TalTech-IRIT-LIS to the DISPLACE 2024 challenge. Our team participated in the speaker diarization and language diarization tracks of the challenge. In the speaker diarization track, our best submission was an ensemble of systems based on the pyannote.audio speaker diarization pipeline utilizing powerset training and our recently proposed PixIT method that performs joint diarization and speech separation. We improve upon PixIT by using the separation outputs for speaker embedding extraction. Our ensemble achieved a diarization error rate of 27.1% on the evaluation dataset. In the language diarization track, we fine-tuned a pre-trained Wav2Vec2-BERT language embedding model on in-domain data, and clustered short segments using AHC and VBx, based on similarity scores from LDA/PLDA. This led to a language diarization error rate of 27.6% on the evaluation data. Both results were ranked first in their respective challenge tracks.
title TalTech-IRIT-LIS Speaker and Language Diarization Systems for DISPLACE 2024
topic Audio and Speech Processing
url https://arxiv.org/abs/2407.12743