Korean Speech Recordings From 211 Mobile Phone Users

Featuring a sample of Korean speech recordings from 211 local speakers, comprising 99 females and 112 males. Audio was captured in a quiet indoor environment using mainstream Android phones and iPhones.

Use Cases

• Train or fine-tune automatic speech recognition models on Korean audio from mobile phone recordings.
• Analyze speech characteristics or build voice activity detection models using audio from a controlled quiet environment.
• Develop speaker verification or gender classification models using data from 211 distinct Korean speakers.

Strengths

• Data from 211 distinct Korean speakers provides a sample of speaker diversity.
• Recordings were made in a quiet indoor environment, controlling for background noise.
• Uses mainstream mobile phones (Android and iPhone) as recording devices, reflecting common audio capture scenarios.

Limitations

• The dataset is described as a sample of a larger paid dataset, so its size and completeness are limited.
• No information is provided on the length, format, or transcription of the audio recordings.
• Demographic details beyond speaker count and gender are not specified.

Provenance

Source

Nexdata

Collection Method

Recorded in a quiet indoor environment using mainstream Android phones and iPhones.

Time Range

null

Freshness

null

Geography

South Korea (based on speaker description as 'Korean locals')