FreedomIntelligence/ExpressiveSpeech

ExpressiveSpeech Dataset Project Webpage 中文版 (Chinese Version) About The Dataset ExpressiveSpeech is a high-quality, expressive, and bilingual (Chinese-English) speech dataset created to address the common lack of consistent vocal expressiveness in existing dialogue datasets. This dataset is meticulously curated from five renowned open-source emotional dialogue datasets: Expresso, NCSSD, M3ED, MultiDialog, and IEMOCAP. Through a rigorous processing and selection… See the full description on the dataset page: https://huggingface.co/datasets/FreedomIntelligence/ExpressiveSpeech.