Hearing-Impaired Speech Database

1. Product Overview

The hearing-impaired speech database comprises 315 speakers and spans 341 hours of audio. It was recorded with native Chinese-speaking participants, at a sampling rate of 16,000 Hz and a bit depth of 16 bits. The database has been transcribed and annotated by professional speech annotators and rigorously quality-checked, achieving a text accuracy rate exceeding 97%.

2. Speech Transcription

Transcribers render the spoken content based on the audio they hear, striving to ensure that the written text accurately reflects the original pronunciation. General guidelines are as follows: 1) The transcription must exactly match the spoken audio—no extra words, missing words, or errors. 2) Annotations must fully capture the actual pronunciation without omissions. 3) Unintelligible segments are marked with an asterisk (*). 4) Whether the speaker uses a level tone is noted, along with recording-specific intonation characteristics. 5) Pronunciation inaccuracies caused by articulation issues are transcribed as heard, while also indicating the type of articulatory error. []: Used to mark repeated complete syllables, multi-syllable units, or words. If only a single sound is repeated, use /r /b: Indicates noticeable prolonged hesitations or brief interruptions. b stands for "block" /p: Marks extended sounds. p stands for "prolongation" /r: Denotes repeated sounds—for example, a single consonant or vowel insufficient to form a full word. r stands for "repetition"  Multiple annotations may be applied to the same word; for instance, if a word is both repeated and elongated, it can be labeled as "I[I/p]".

3. Speaker Information

The database includes a total of 315 speakers—165 males and 150 females.

Gender · Male · Female · Total

Percentage · 52.4% · 47.6% · 100%

· Age Group · Number of Speakers · Percentage

A · 16–17 years old · 14 · 4.45%

B · 18–22 years old · 235 · 74.60%

C · 23–30 years old · 66 · 20.95%

Total · 315 · 100% ·

Age Distribution

The specific age distribution of the speakers is as follows: A (16–17 years); B (18–22 years); C (23–30 years).

4. Directory Structure and Naming Conventions

Directory Structure

Data directory tree structure:

└─DOC (Textual Documentation)
││├─content.txt (List of Transcribed Content)
││├─spker-info.xlsx (Speaker Information Table)
└─SPEECHDATA (Audio Folder)
│├─0095 (Speaker ID)
││├─TL095_001.wav (Audio File)

Naming Conventions

CORPUS/USAGE/SPEAKER/SPEECH_IDe.g./SPEECHDATA/0095/TL0095_001.wav

Directory Name · Content · Notes

CORPUS · · Speech Database Identifier

USAGE · SPEECHDATA · Audio Folder

SPEAKER · 0095 · Speaker ID

SPEECH_ID · TL0095_001.wav · Audio File

Sample Data Download

Click to download sample data (data_sample.zip)