David A. van Leeuwen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
David A. van Leeuwen has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Henk van Den Heuvel
Centre for Language and Speech Technology, Radboud University, Nijmegen, The Netherlands
5 shared publications - Mitchell McLaren
SRI International, Menlo Park, CA, USA
3 shared publications - Emre Yılmaz
CLS/CLST, Radboud University, Nijmegen, Netherlands
2 shared publications Albert Swart
AGNITIO Research, South Africa + Radboud University, Nijmegen, The Netherlands
1 shared publicationEmre Yılmaz
CLS/CLST Radboud University Nijmegen Netherlands
1 shared publication- Emre Yılmaz
CLS/CLST Radboud University Erasmusplein 1 6525 HT Nijmegen Netherlands
1 shared publication
Latest Publications
- 15 December 2019
- Published by John Benjamins Publishing Company
- p. 86-101
Chapter 5. Language change caught in the act
This study investigates language change in Frisian relative pronouns in a corpus of Frisian radio broadcasts (1966–2015). In spite of the limitations of this corpus, we were able to catch language change in the act. The analyses show that until the 1980s the younger speakers in these broadcasts lead the rise of t-full relative pronouns, a change that was first observed in literature at the end of the 19th century. From the 1980s onwards the new younger generation reversed this change and increasingly started using the t-less relative pronouns. Additionally, the increase of t-less forms occurred mostly in non-scripted, spontaneous speech. It does not seem to play a role whether the speaker is a presenter or an interviewee/guest.
- 15 December 2019
- Published by John Benjamins Publishing Company
- p. 86-101
Chapter 5. Language change caught in the act
This study investigates language change in Frisian relative pronouns in a corpus of Frisian radio broadcasts (1966–2015). In spite of the limitations of this corpus, we were able to catch language change in the act. The analyses show that until the 1980s the younger speakers in these broadcasts lead the rise of t-full relative pronouns, a change that was first observed in literature at the end of the 19th century. From the 1980s onwards the new younger generation reversed this change and increasingly started using the t-less relative pronouns. Additionally, the increase of t-less forms occurred mostly in non-scripted, spontaneous speech. It does not seem to play a role whether the speaker is a presenter or an interviewee/guest.
- 1 December 2018
- Published by Elsevier in Speech Communication
- Vol. 105, 12-22
Semi-supervised acoustic model training for speech with code-switching
- 1 December 2018
- Published by Elsevier in Speech Communication
- Vol. 105, 12-22
Semi-supervised acoustic model training for speech with code-switching
- Open Access
Semi-supervised acoustic model training for speech with code-switching
In the FAME! project, we aim to develop an automatic speech recognition (ASR) system for Frisian-Dutch code-switching (CS) speech extracted from the archives of a local broadcaster with the ultimate goal of building a spoken document retrieval system. Unlike Dutch, Frisian is a low-resourced language with a very limited amount of manually annotated speech data. In this paper, we describe several automatic annotation approaches to enable using of a large amount of raw bilingual broadcast data for acoustic model training in a semi-supervised setting. Previously, it has been shown that the best-performing ASR system is obtained by two-stage multilingual deep neural network (DNN) training using 11 hours of manually annotated CS speech (reference) data together with speech data from other high-resourced languages. We compare the quality of transcriptions provided by this bilingual ASR system with several other approaches that use a language recognition system for assigning language labels to raw speech segments at the front-end and using monolingual ASR resources for transcription. We further investigate automatic annotation of the speakers appearing in the raw broadcast data by first labeling with (pseudo) speaker tags using a speaker diarization system and then linking to the known speakers appearing in the reference data using a speaker recognition system. These speaker labels are essential for speaker-adaptive training in the proposed setting. We train acoustic models using the manually and automatically annotated data and run recognition experiments on the development and test data of the FAME! speech corpus to quantify the quality of the automatic annotations. The ASR and CS detection results demonstrate the potential of using automatic language and speaker tagging in semi-supervised bilingual acoustic model training.
Open AccessSemi-supervised acoustic model training for speech with code-switching
In the FAME! project, we aim to develop an automatic speech recognition (ASR) system for Frisian-Dutch code-switching (CS) speech extracted from the archives of a local broadcaster with the ultimate goal of building a spoken document retrieval system. Unlike Dutch, Frisian is a low-resourced language with a very limited amount of manually annotated speech data. In this paper, we describe several automatic annotation approaches to enable using of a large amount of raw bilingual broadcast data for acoustic model training in a semi-supervised setting. Previously, it has been shown that the best-performing ASR system is obtained by two-stage multilingual deep neural network (DNN) training using 11 hours of manually annotated CS speech (reference) data together with speech data from other high-resourced languages. We compare the quality of transcriptions provided by this bilingual ASR system with several other approaches that use a language recognition system for assigning language labels to raw speech segments at the front-end and using monolingual ASR resources for transcription. We further investigate automatic annotation of the speakers appearing in the raw broadcast data by first labeling with (pseudo) speaker tags using a speaker diarization system and then linking to the known speakers appearing in the reference data using a speaker recognition system. These speaker labels are essential for speaker-adaptive training in the proposed setting. We train acoustic models using the manually and automatically annotated data and run recognition experiments on the development and test data of the FAME! speech corpus to quantify the quality of the automatic annotations. The ASR and CS detection results demonstrate the potential of using automatic language and speaker tagging in semi-supervised bilingual acoustic model training.
- Open Access
Code-Switching Detection with Data-Augmented Acoustic and Language Models
In this paper, we investigate the code-switching detection performance of a code-switching (CS) automatic speech recognition (ASR) system with data-augmented acoustic and language models. We focus on the recognition of Frisian-Dutch radio broadcasts where one of the mixed languages, namely Frisian, is under-resourced. Recently, we have explored how the acoustic modeling (AM) can benefit from monolingual speech data belonging to the high-resourced mixed language. For this purpose, we have trained state-of-the-art AMs on a significantly increased amount of CS speech by applying automatic transcription and monolingual Dutch speech. Moreover, we have improved the language model (LM) by creating CS text in various ways including text generation using recurrent LMs trained on existing CS text. Motivated by the significantly improved CS ASR performance, we delve into the CS detection performance of the same ASR system in this work by reporting CS detection accuracies together with a detailed detection error analysis.
Open AccessCode-Switching Detection with Data-Augmented Acoustic and Language Models
In this paper, we investigate the code-switching detection performance of a code-switching (CS) automatic speech recognition (ASR) system with data-augmented acoustic and language models. We focus on the recognition of Frisian-Dutch radio broadcasts where one of the mixed languages, namely Frisian, is under-resourced. Recently, we have explored how the acoustic modeling (AM) can benefit from monolingual speech data belonging to the high-resourced mixed language. For this purpose, we have trained state-of-the-art AMs on a significantly increased amount of CS speech by applying automatic transcription and monolingual Dutch speech. Moreover, we have improved the language model (LM) by creating CS text in various ways including text generation using recurrent LMs trained on existing CS text. Motivated by the significantly improved CS ASR performance, we delve into the CS detection performance of the same ASR system in this work by reporting CS detection accuracies together with a detailed detection error analysis.
- Open Access
- 1 December 2017
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 142 (6) , EL519-EL524
Phonetic similarity of /s/ in native and second language: Individual differences in learning curves
Second language learners may merge similar sounds from their native (L1) and second (L2) languages into a single phonetic category, neutralizing subphonemic differences in these similar sounds. This study investigates whether Dutch speakers produce phonetically distinct variants of /s/ in their L1 Dutch and L2 English, and whether and how this phonetic categorization develops over time. Target /s/ sounds in matching words in L1 and L2 were compared in their centre of spectral gravity. Speakers varied in their individual learning curves in the categorization of produced /s/ sounds, both in starting points and in longitudinal trajectories. After 3 years, however, all speakers had converged in producing their /s/ variants in L1 and L2 as two similar but different sounds.
- 1 December 2017
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 142 (6) , EL519-EL524
Open AccessPhonetic similarity of /s/ in native and second language: Individual differences in learning curves
Second language learners may merge similar sounds from their native (L1) and second (L2) languages into a single phonetic category, neutralizing subphonemic differences in these similar sounds. This study investigates whether Dutch speakers produce phonetically distinct variants of /s/ in their L1 Dutch and L2 English, and whether and how this phonetic categorization develops over time. Target /s/ sounds in matching words in L1 and L2 were compared in their centre of spectral gravity. Speakers varied in their individual learning curves in the categorization of produced /s/ sounds, both in starting points and in longitudinal trajectories. After 3 years, however, all speakers had converged in producing their /s/ variants in L1 and L2 as two similar but different sounds.
- Open Access
The "Sprekend Nederland" project and its application to accent location
This paper describes the data collection effort that is part of the project Sprekend Nederland (The Netherlands Talking), and discusses its potential use in Automatic Accent Location. We define Automatic Accent Location as the task to describe the accent of a speaker in terms of the location of the speaker and its history. We discuss possible ways of describing accent location, the consequence these have for the task of automatic accent location, and potential evaluation metrics.
Open AccessThe "Sprekend Nederland" project and its application to accent location
This paper describes the data collection effort that is part of the project Sprekend Nederland (The Netherlands Talking), and discusses its potential use in Automatic Accent Location. We define Automatic Accent Location as the task to describe the accent of a speaker in terms of the location of the speaker and its history. We discuss possible ways of describing accent location, the consequence these have for the task of automatic accent location, and potential evaluation metrics.
- Open Access
- 16 June 2014
- Published by International Speech Communication Association
- p. 14-18
A comparison of linear and non-linear calibrations for speaker recognition
- 16 June 2014
- Published by International Speech Communication Association
- p. 14-18
Open AccessA comparison of linear and non-linear calibrations for speaker recognition
- 22 August 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech, and Language Processing
- Vol. 21 (11) , 2425-2438
Quality Measure Functions for Calibration of Speaker Recognition Systems in Various Duration Conditions
This paper investigates the effect of utterance duration to the calibration of a modern i-vector speaker recognition system with probabilistic linear discriminant analysis (PLDA) modeling. A calibration approach to deal with these effects using quality measure functions (QMFs) is proposed to include duration in the calibration transformation. Extensive experiments are performed in order to evaluate the robustness of the proposed calibration approach for unseen conditions in the training of calibration parameters. Using the latest NIST corpora for evaluation, results highlight the importance of considering the quality metrics like duration in calibrating the scores for automatic speaker recognition systems.
- 22 August 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Audio, Speech, and Language Processing
- Vol. 21 (11) , 2425-2438
Quality Measure Functions for Calibration of Speaker Recognition Systems in Various Duration Conditions
This paper investigates the effect of utterance duration to the calibration of a modern i-vector speaker recognition system with probabilistic linear discriminant analysis (PLDA) modeling. A calibration approach to deal with these effects using quality measure functions (QMFs) is proposed to include duration in the calibration transformation. Extensive experiments are performed in order to evaluate the robustness of the proposed calibration approach for unseen conditions in the training of calibration parameters. Using the latest NIST corpora for evaluation, results highlight the importance of considering the quality metrics like duration in calibrating the scores for automatic speaker recognition systems.
- Open Access
- 11 November 2012
- Published by Springer Nature
- p. 271-288
N-Best 2008: A Benchmark Evaluation for Large Vocabulary Speech Recognition in Dutch
- 11 November 2012
- Published by Springer Nature
- p. 271-288
Open AccessN-Best 2008: A Benchmark Evaluation for Large Vocabulary Speech Recognition in Dutch
- 1 November 2012
- Published by Elsevier in Speech Communication
- Vol. 54 (9) , 1049-1063
Speech-based recognition of self-reported and observed emotion in a dimensional space
- 1 November 2012
- Published by Elsevier in Speech Communication
- Vol. 54 (9) , 1049-1063
Speech-based recognition of self-reported and observed emotion in a dimensional space