Kinyarwanda Speech Recordings from Studio Voice Actress

A collection of 3,992 audio clips of Kinyarwanda text-to-speech recordings made by a single voice actress in a studio setting. It was collected as part of the Mbaza project and includes a CSV file linking audio file names to their corresponding written text.

Use Cases

• Train a text-to-speech model on 3,992 studio-quality audio clips paired with Kinyarwanda text.
• Fine-tune a speech synthesis model using the single female voice actress recordings for consistent vocal characteristics.
• Analyze the phonetic coverage of the Kinyarwanda syllabes included in the collected text data.

Strengths

• 3,992 audio clips provide a foundation for model training.
• Studio recordings by a single voice actress ensure consistent audio quality and vocal characteristics.
• Includes a CSV file linking audio names to their corresponding text for alignment.

Limitations

• Dataset size of 3,992 clips is relatively small for training robust, generalizable TTS models.
• Limited to a single voice actress, which restricts vocal diversity and accent representation.
• Specific details on audio format, duration, and text complexity are not provided in the input.

Provenance

Source

mbazaNLP on Hugging Face.

Collection Method

Studio recordings by a voice actress, with text collected to include Kinyarwanda syllabes.

Freshness

Last updated on 2023-06-27.