Vietnamese Speech Audio With Text Transcripts

Featuring 9.5 hours of Vietnamese speech audio paired with text transcripts, totaling 1.28GB. The audio was crawled from YouTube audiobooks, and the text was labeled by VinBrain JSC.

Use Cases

• Train a text-to-speech model using the paired audio and text samples.
• Fine-tune an automatic speech recognition model on Vietnamese audio with labeled text transcripts.
• Analyze prosody and pronunciation patterns in Vietnamese audiobook speech from the provided audio-text pairs.

Strengths

• 9.5 hours of total audio duration provides a substantial corpus for model training.
• Text transcripts are professionally labeled by VinBrain JSC, suggesting quality annotations.
• Dataset size of 1.28GB indicates a significant volume of audio data.

Limitations

• The audio is sourced from YouTube audiobooks, which may introduce variable recording quality and background noise.
• The dataset's temporal coverage and geographic representativeness within Vietnam are unknown.
• The specific number of audio-text pairs (rows) is not provided, making scale assessment difficult.

Provenance

Source

Audio crawled from YouTube audiobooks; text labeled by VinBrain JSC.

Collection Method

Audio crawled from YouTube; text labeling.

Time Range

null

Freshness

null

Geography

Vietnam (inferred from language and author).