MagicData

sign in

Total Size: 5.4 GB

Dataset Overview

Dataset Type

ASR speech corpus

Language

zh-CN,
Mandarin Chinese (China)

Speech Style

spontaneous conversation

Content

themed conversations

Audio Parameters

16 kHz, 16 bits,
6 channels

File Format

WAV (PCM)
TXT (UTF-8)

Recording Equipment

microphone, mobile, and recorder

Recording Environment

indoor environment

License

Magic Data
open-source license

Open Source
ASR Corpus
10 hours

ASR-MultiDeviCCSC: A Chinese Conversational Speech Corpus

10 hours of transcribed Mandarin Chinese conversational speech
collected by multiple devices

This open-source dataset consists of 10 hours of transcribed Mandarin Chinese conversational speech on certain topics, including six-channel audios collected by three devices, where 360 conversations contributed by 30 pairs of speakers were contained.

Sample for iOS Phone:

Sample for Android Phone:

Sample for Recorder:

Sample for Microphone:

Dataset Overview

Dataset Type

ASR speech corpus

Language

zh-CN,
Mandarin Chinese (China)

Speech Style

spontaneous conversation

Content

themed conversations

Audio Parameters

16 kHz, 16 bits,
6 channels

File Format

WAV (PCM)
TXT (UTF-8)

Recording Equipment

microphone, mobile, and recorder

Recording Environment

indoor environment

License

Magic Data
open-source license

{{ reviewsTotal }}{{ options.labels.singularReviewCountLabel }}
{{ reviewsTotal }}{{ options.labels.pluralReviewCountLabel }}
{{ options.labels.newReviewButton }}
{{ userData.canReview.message }}

Verifying Email