Stuttering Detection Using Speaker Representations and Self-supervised Contextual Embeddings - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Article Dans Une Revue International Journal of Speech Technology Année : 2023

Stuttering Detection Using Speaker Representations and Self-supervised Contextual Embeddings

Résumé

The adoption of advanced deep learning architectures in stuttering detection (SD) tasks is challenging due to the limited size of the available datasets. To this end, this work introduces the application of speech embeddings extracted from pre-trained deep learning models trained on large audio datasets for different tasks. In particular, we explore audio representations obtained using emphasized channel attention, propagation, and aggregation time delay neural network (ECAPA-TDNN) and Wav2Vec2.0 models trained on VoxCeleb and LibriSpeech datasets respectively. After extracting the embed- dings, we benchmark with several traditional classifiers, such as the K-nearest neighbour (KNN), Gaussian naive Bayes, and neural network, for the SD tasks. In comparison to the standard SD systems trained only on the limited SEP-28k dataset, we obtain a relative improvement of 12.08%, 28.71%, 37.9% in terms of unweighted average recall (UAR) over the baselines. Finally, we have shown that combining two embeddings and concatenating multiple layers of Wav2Vec2.0 can further improve the UAR by up to 2.60% and 6.32% respectively.
Fichier principal
Vignette du fichier
paper (1).pdf (5.1 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03629758 , version 1 (04-04-2022)
hal-03629758 , version 2 (02-06-2023)

Licence

Paternité

Identifiants

Citer

Shakeel Sheikh, Md Sahidullah, Fabrice Hirsch, Slim Ouni. Stuttering Detection Using Speaker Representations and Self-supervised Contextual Embeddings. International Journal of Speech Technology, 2023, ⟨10.1007/s10772-023-10032-1⟩. ⟨hal-03629758v2⟩
302 Consultations
298 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More