HUI-Audio-Corpus-German: A high quality TTS dataset

Abstract

The increasing availability of audio data on the internet lead to a multitude of datasets for development and training of text to speech applications, based on neural networks. Highly differing quality of voice, low sampling rates, lack of text normalization and disadvantageous alignment of audio samples to corresponding transcript sentences still limit the performance of deep neural networks trained on this task. Additionally, data resources in languages like German are still very limited. We introduce the "HUI-Audio-Corpus-German", a large, open-source dataset for TTS engines, created with a processing pipeline, which produces high quality audio to transcription alignments and decreases manual effort needed for creation. mehr

Mehr zum Titel

Titel HUI-Audio-Corpus-German: A high quality TTS dataset
Medien 44th German Conference on Artificial Intelligence (KI2021)
Verlag ---
Heft ---
Band ---
ISBN ---
Verfasser/Herausgeber Pascal Puchtler, Johannes Wirth, Prof. Dr. René Peinl
Seiten ---
Veröffentlichungsdatum 11.06.2021
Projekttitel ---
Zitation Puchtler, Pascal; Wirth, Johannes; Peinl, René (2021): HUI-Audio-Corpus-German: A high quality TTS dataset. 44th German Conference on Artificial Intelligence (KI2021).