Enhanced CORILGA: Introducing the Automatic Phonetic Alignment Tool for Continuous Speech

Enhanced CORILGA: Introducing the Automatic Phonetic Alignment Tool for Continuous Speech Roberto Seara*, Marta Martinez*, Rocío Varela*, Carmen García-Mateo*, Elisa Fernández-Rei**, Xosé Luis Regueira** * AtlantTIC Research Center - Escola de Enxeñaría de Telecomunicación- Universidade de Vigo Campus Universitario 36310 Vigo (Spain) ** Instituto da Lingua Galega Universidade de Santiago de Compostela Praza da Universidade, 4, 15782 Santiago de Compostela (Spain) E-mail: carmen.garcia@uvigo.es Abstract The Corpus Oral Informatizado da Lingua Galega (CORILGA) project aims at building a corpus of oral language for Galician, primarily designed to study the linguistic variation and change. This project is currently under development and it is periodically enriched with new contributions. The long-term goal is that all the speech recordings will be enriched with phonetic, syllabic, morphosyntactic, lexical and sentence ELAN-complaint annotations. A way to speed up the process of annotation is to use automatic speech-recognition-based tools tailored to the application. Therefore, CORILGA repository has been enhanced with an automatic alignment tool, available to the administrator of the repository, that aligns speech with an orthographic transcription. In the event that no transcription, or just a partial one, were available, a speech recognizer for Galician is used to generate word and phonetic segmentations. These recognized outputs may contain errors that will have to be manually corrected by the administrator. For assisting this task, the tool also provides an ELAN tier with the confidence measure of each recognized word. In this paper, after the description of the main facts of the CORILGA corpus, the speech alignment and recognition tools are described. Both have been developed using the Kaldi toolkit. Keywords: minority languages, alignment and recognition tool, confidence measure 1. Introduction The Corpus Oral Informatizado da Lingua Galega (CORILGA) (Garcia-Mateo, 2014) is a corpus of oral language for Galician language that was designed to study the linguistic variation and change, but it can also be useful for other purposes (language teaching, linguistic repertoires, etc.). The corpus consists of spoken language recordings of different levels and registers (standard language, literary, popular, formal and colloquial language, rural and urban) and of different types: conversations, guided interviews, speeches, lectures, recited poetry, theatre and media. It collects the language of speakers of different ages and both sexes that were recorded from mid-1960 to the present time. Apart from the recordings specifically made for this corpus and the public oral texts taken from the media or the Internet, CORILGA incorporates texts and, in some cases, transcriptions from other corpora, among which they are the projects of the Arquivo do Galego Oral (AGO) (Fernández Rei, 2011), Prosodia da lingua formal, AMPER-Galicia (Escourido, 2008), or the personal corpora, namely those from Gustav Henningsen (Vázquez, 2012), from Francisco Dubert about Santiago de Compostela, from Xosé Luís Regueira about Vilalba region, Manuel Rico including radio interviews, from Noemi Basanta including conversations, from Eduardo Louredo about Leiro, from Miguel Abraira and Xabier Iglesias. The corpus is currently under development, and it is periodically enriched with new contributions, either recordings, or aligned and annotated texts. For this reason, not …

Content truncated to 500 words