The Influence of Dataset Partitioning on Dysfluency Detection Systems

Bayerl SP, Wagner D, Nöth E, Bocklet T, Riedhammer K (2022)

Publication Type: Conference contribution

Publication year: 2022

Journal

Lecture Notes in Computer Science Springer Verlag

Publisher: Springer Science and Business Media Deutschland GmbH

Book Volume: 13502 LNAI

Pages Range: 423-436

Conference Proceedings Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Event location: Brno CZ

ISBN: 9783031162695

DOI: 10.1007/978-3-031-16270-1_35

Abstract

This paper empirically investigates the influence of different data splits and splitting strategies on the performance of dysfluency detection systems. For this, we perform experiments using wav2vec 2.0 models with a classification head as well as support vector machines (SVM) in conjunction with the features extracted from the wav2vec 2.0 model to detect dysfluencies. We train and evaluate the systems with different non-speaker-exclusive and speaker-exclusive splits of the Stuttering Events in Podcasts (SEP-28k) dataset to shed some light on the variability of results w.r.t. to the partition method used. Furthermore, we show that the SEP-28k dataset is dominated by only a few speakers, making it difficult to evaluate. To remedy this problem, we created SEP-28k-Extended (SEP-28k-E), containing semi-automatically generated speaker and gender information for the SEP-28k corpus, and suggest different data splits, each useful for evaluating other aspects of methods for dysfluency detection.

Authors with CRIS profile

Elmar Nöth Professur für Informatik (Mustererkennung) Tobias Bocklet Professur für Informatik (Mustererkennung)

Involved external institutions

Technische Hochschule Nürnberg "Georg Simon Ohm" DE Germany (DE)

How to cite

APA:

Bayerl, S.P., Wagner, D., Nöth, E., Bocklet, T., & Riedhammer, K. (2022). The Influence of Dataset Partitioning on Dysfluency Detection Systems. In Petr Sojka, Aleš Horák, Ivan Kopeček, Karel Pala (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 423-436). Brno, CZ: Springer Science and Business Media Deutschland GmbH.

MLA:

Bayerl, Sebastian P., et al. "The Influence of Dataset Partitioning on Dysfluency Detection Systems." Proceedings of the 25th International Conference on Text, Speech, and Dialogue, TSD 2022, Brno Ed. Petr Sojka, Aleš Horák, Ivan Kopeček, Karel Pala, Springer Science and Business Media Deutschland GmbH, 2022. 423-436.

BibTeX: Download