Taiwan Mandarin Speech Recordings from 204 Mobile Phone Readers

Comprising a sample of Taiwan Mandarin speech recordings from 204 residents, each reading 450 sentences. The content covers topics like economy, entertainment, news, and spoken language, including general and human-computer interaction scenes. The audio data is accompanied by manual transcriptions.

Use Cases

• Train an automatic speech recognition (ASR) model on manually transcribed Taiwan Mandarin audio recordings.
• Analyze phonetic and prosodic variations in Taiwan Mandarin across 204 speakers reading 450 sentences each.
• Develop voice assistants or dialogue systems using speech data covering human-computer interaction scenes and general topics.

Strengths

• Data from 204 distinct Taiwan resident speakers provides speaker diversity.
• Each speaker contributed 450 sentences, resulting in a substantial number of utterances per speaker.
• Content covers multiple domains including economy, entertainment, and news, offering topical variety.

Limitations

• The dataset is described as a sample of a larger paid dataset, so its full scale is unknown.
• Specific audio quality metrics, recording conditions, and speaker demographic details are not provided.
• The availability of validation/test splits or accompanying metadata is unclear from the description.

Provenance

Source

Nexdata

Collection Method

Recorded via mobile phone reading by 204 Taiwan residents.

Freshness

The dataset was last updated on 2025-04-23.

Geography

Taiwan