PEASYV: a procedure to obtain phonetic data from subtitled videos - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

PEASYV: a procedure to obtain phonetic data from subtitled videos

Résumé

This paper presents a procedure to easily obtain phonetic data from subtitled YouTube videos in English. Called PEASYV (Phonetic Extraction and Alignment of Subtitled YouTube Videos) and based on a UNIX shell script, this procedure makes it possible to obtain fine-grained phonetic and phonological data from any subtitled videos posted on YouTube. It uses two aligners, P2FA [1] and SPPAS [2], to generate Praat [3] TextGrids for each video. These TextGrids contain tiers specific to each aligner, with segmental, syllabic and lexical alignments. SPPAS tiers also contain prosodic information. After describing the procedure and the generated files, a case study of a YouTube channel on English pronunciation is presented, in order to assess the quality of the alignment: the formants of the monophthongs from the channel's 452 videos, amounting to more than 179 hours of spontaneous speech, are analysed. The potential alignment errors, along with their prevention, are then discussed.
Fichier principal
Vignette du fichier
ICPhS2023MeliBallier.pdf (151.49 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte
Licence : CC BY - Paternité

Dates et versions

hal-04319467 , version 1 (02-12-2023)

Licence

Paternité

Identifiants

  • HAL Id : hal-04319467 , version 1

Citer

Nicolas Ballier, Adrien Méli. PEASYV: a procedure to obtain phonetic data from subtitled videos. 20th International Congress of Phonetic Sciences, Aug 2023, Prague, Czech Republic. pp.3211-3215. ⟨hal-04319467⟩
42 Consultations
18 Téléchargements

Partager

Gmail Facebook X LinkedIn More