## About the Dataset
### 382 hours

This audio dataset contains 382 hours of Mandarin Chinese Speech Data in generic domain.

The speakers are presented with a prompt (script) and asked to read it out loud and record. Our clients will receive an audio recording, the prompt and information about the speaker. The audio is recorded on-device, typically in 16Khz 16 bit. We also provide information on which device each record was recorded.

The dataset is covered by [Defined.ai's standard license agreement](https://www.defined.ai/dataset/data-licence-agreement). The license agreement is perpetual and allows for the commercialization of all models built on the data.



Request Sample

Spontaneous IVR

Mandarin Chinese Spontaneous IVR

Speech

Mandarin Chinese (PRC) Scripted Monologue 

Korean Spontaneous Dialogue

Japanese Spontaneous Dialogue

You might also be interested in these audio datasets:

## About the Dataset
### 48 hours

This audio dataset contains 48 hours of Mandarin Chinese Speech Data in various domains, recorded by speakers from PRC.

Domain distribution per dataset:

- 12.12 hours of Banking
- 11.88 hours of Insurance
- 11.93 hours of Retail
- 12.4 hours of Telecommunication

IVR stands for Interactive Voice Response. The Spontaneous IVR data is created by having a human respond to an IVR system. The human (playing a customer) is following simplified real-life scenarios, making a spontaneous query on the given topic. The IVR system will ask the human to repeat his/her query in 2 more different ways. The speech is then transcribed. In rare cases, the IVR system is mimicked by another human. The recording is done via telephony and is saved in 8khz 16 bit per channel.

The dataset is covered by [Defined.ai's standard license agreement](https://www.defined.ai/dataset/data-licence-agreement). The license agreement is perpetual and allows for the commercialization of all models built on the data.

Other characteristics:

- Audio format: WAV
- Recording environment: noisy, silent
- Communication band: narrowband
- Sample rate: 8Hz
- Device type: mobile

## Samples

Check this 5-minute audio sample [here](https://defineddata.blob.core.windows.net/samples/AFI_zh-cn_single-spontaneous-IVR_banking_30m_v01_Sample/Audio/ba474f34-275b-4a65-8948-d25d341e2825.wav?se=2024-06-15T17%3A05%3A09Z&sp=r&sv=2020-06-12&ss=b&srt=o&sig=iWjw%2BCnLPJth9mcE63Zg70m4v4gO8tqvnrSWzFK9MNQ%3D). Transcription for the sample is also [available](https://prdstrapimediastorage.blob.core.windows.net/prdstrapimediastorage/assets/IVR_Chinese_Banking_Short_Transcription_9c13870ae0.tsv)

## About the Dataset
### 422 hours

This audio dataset contains 422 hours of Japanese Speech Data in various domains, recorded by speakers from Japan.

Domain distribution per dataset:

- 105.17 hours of Banking
- 103.28 hours of Insurance
- 107.48 hours of Retail
- 106.27 hours of Telecommunication

Defined.ai creates scenarios for our crowd members to follow, which they study beforehand. They then record a conversation, one speaker playing the agent, the other speaker “playing out” the scenario with spontaneous content. The recording is done via telephony and is saved in 8kHz 16 bit per channel. That content is then transcribed.

The dataset is covered by [Defined.ai's standard license agreement](https://www.defined.ai/dataset/data-license-agreement). The license agreement is perpetual and allows for the commercialization of all models built on the data.

Other characteristics:

- Audio format: WAV
- Recording environment: noisy, silent
- Bits per sample: 16
- Communication band: broadband
- Sample rate: 8kHz

## Metadata Distribution

![Spontaneous_Japanese_Gender.png](https://prdstrapimediastorage.blob.core.windows.net/prdstrapimediastorage/assets/Spontaneous_Japanese_Gender_be95c02e78.png)
![Spontaneous_Japanese_Age.png](https://prdstrapimediastorage.blob.core.windows.net/prdstrapimediastorage/assets/Spontaneous_Japanese_Age_8c011760f7.png)

## Samples

Check this 5-minute audio sample [here](https://defineddata.blob.core.windows.net/samples/AFI_ja-jp_multi-spontaneous_insurance_30m_v01_Sample/Audio/dac3886b-94b8-4839-a1b2-b7da1f716c16.wav?se=2024-06-15T15%3A43%3A57Z&sp=r&sv=2020-06-12&ss=b&srt=o&sig=fvTybGA394qfxK/jKA68Dei7fAYwl2ZuriLgt/wfrCs%3D). Transcription for the sample is also [available](https://prdstrapimediastorage.blob.core.windows.net/prdstrapimediastorage/assets/Spontaneous_Japanese_Insurance_Short_Transcription_b7d9f1cc9f.tsv)

## About the Dataset
### 521 hours

This audio dataset contains 521 hours of Korean Speech Data in various domains, recorded by speakers from Korea.

Domain distribution per dataset:

- 147.83 hours of Banking
- 147.92 hours of Insurance
- 111.9 hours of Retail
- 113.85 hours of Telecommunication

Defined.ai creates scenarios for our crowd members to follow, which they study beforehand. They then record a conversation, one speaker playing the agent, the other speaker “playing out” the scenario with spontaneous content. The recording is done via telephony and is saved in 8khz 16 bit per channel. That content is then transcribed.

The dataset is covered by [Defined.ai's standard license agreement](https://www.defined.ai/dataset/data-licence-agreement). The license agreement is perpetual and allows for the commercialization of all models built on the data.

Other characteristics:

- Audio format: WAV
- Recording environment: noisy, silent
- Bits per sample: 16
- Communication band: broadband
- Sample rate: 8Hz

## Samples

Check this 5-minute audio sample [here](https://defineddata.blob.core.windows.net/samples/AFI_ko-ko_multi-spontaneous_telco_B1_v01_30m_Sample/Audio/740b25f2-aadb-4f06-b608-abd52d9f2783.wav?se=2025-07-06T12%3A46%3A26Z&sp=r&sv=2020-06-12&ss=b&srt=o&sig=PxfKSWYG1bu/apOjH21OzeEeNWs%2BuWzXemj/oUWXKDM%3D). Transcription for the sample is also [available](https://prdstrapimediastorage.blob.core.windows.net/prdstrapimediastorage/assets/Spontaneous_Korean_Telco_Short_Transcription_229b4ba4b7.tsv)