Model-generated Chinese emotional synthetic speech database
About this dataset
1. Product Overview
The model-generated Chinese emotional synthetic speech database contains a total of 462,000 entries, collected from 92 speakers. The synthesis engines used include iFLYTEK, Volcano, Alibaba, and Microsoft: iFLYTEK (16 kHz, 16-bit); Volcano (24 kHz, 16-bit); Alibaba (16 kHz/24 kHz/48 kHz, 16-bit); and Microsoft (48 kHz, 16-bit).
2. Product Directory Structure
Directory Structure
Data directory structure
└─spker-info.xlsx (voice tone information)
└─SPEECHDATA (data folder)
│├─XF (synthesis platform)
│├─GuanShan (voice tone name)
│├─Standard (emotion/style classification)
││ ││ 0000001.pcm (audio file)
││ ││ 0000001.txt (speech text content)
│├─HS (synthesis platform)
Naming Convention
CORPUS/USAGE/PLATFORM/SPEAKER/STYLE/SPEECH_IDe.g./SPEECHDATA/XF/GuanShan/Standard/0000001.pcm
Directory Name · Content · Notes
CORPUS · · Speech database ID
USAGE · SPEECHDATA · Folder name
PLATFORM · XF · Synthesis interface platform
SPEAKER · GuanShan · Voice tone name
STYLE · Standard · Style or emotion category
SPEECH_ID 0000001.pcm WAV file
3. Speaker Information
The gender distribution of speakers is shown in the table below, with a total of 92 voice tones covering various platforms, plus an additional 30 emotion styles. Each speaker contributes 2,000 entries per emotion style, with each utterance lasting at least 5 seconds.
Gender · Number of Speakers · Percentage
Male · 31 · 33.7%
Female · 54 · 58.7%
Children · 7 · 7.6%
Total · 92 · 100%
Sample Data Download
Samples
- Audio sample 1
- Audio sample 2
- Audio sample 3