VoiceHome-2, an extended corpus for multichannel speech processing in real homes - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Speech Communication Année : 2019

VoiceHome-2, an extended corpus for multichannel speech processing in real homes

Résumé

We present a new, extended version of the voiceHome corpus for distant-microphone speech processing in domestic environments. This 5-hour corpus includes short reverberated, noisy utterances (smart home commands) spoken in French by 12 native French talkers in diverse realistic acoustic conditions and recorded by an 8-microphone device at various angles and distances and in various noise conditions. Noise-only segments before and after each utterance are included in the recordings. Clean speech and spontaneous speech recorded in 12 real rooms distributed in 4 different homes are also available. All data have been fully annotated. At last, we provide baseline software for speaker and noise localization, enhancement by source separation, and automatic speech recognition. This corpus stands apart from other corpora in the field by the number of rooms and homes considered and by the fact that it is publicly available at no cost. We describe the corpus specifications and annotations and the data recorded so far, and we report baseline results.
Fichier principal
Vignette du fichier
bertin_SpeechCom18.pdf (909.05 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01923108 , version 1 (15-11-2018)

Identifiants

Citer

Nancy Bertin, Ewen Camberlein, Romain Lebarbenchon, Emmanuel Vincent, Sunit Sivasankaran, et al.. VoiceHome-2, an extended corpus for multichannel speech processing in real homes. Speech Communication, 2019, 106, pp.68-78. ⟨10.1016/j.specom.2018.11.002⟩. ⟨hal-01923108⟩
387 Consultations
723 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More