French Speech Audio Samples With Manual Transcripts

Encompassing 231 hours of French speech audio recorded by 406 speakers from France, Canada, and Africa. The audio was recorded in a quiet environment via mobile phone reading and covers content from fields like economics, entertainment, and news. All audio has corresponding manually transcribed text with a reported sentence accuracy rate of 95%.

Use Cases

• Train an automatic speech recognition (ASR) model on manually transcribed French audio recordings.
• Analyze regional pronunciation variations using audio from speakers in France, Canada, and Africa.
• Build a text-to-speech (TTS) model or evaluate speech synthesis using high-quality, quiet-environment recordings.
• Perform content classification on transcribed text across economics, entertainment, and news domains.

Strengths

• 231 hours of audio data provides a substantial sample for speech-related tasks.
• 406 speakers contribute to speaker diversity, covering multiple French-speaking regions.
• Manual transcription with a 95% sentence accuracy rate indicates high-quality labels.
• Recordings were made in a quiet environment, reducing background noise.

Limitations

• The dataset is described as a sample, suggesting it is not the complete paid dataset.
• Specific audio file formats, sampling rates, and metadata columns are not provided.
• The geographic distribution of the 406 speakers across the three regions is not detailed.

Provenance

Source

Nexdata

Collection Method

Recorded by speakers reading text on mobile phones in a quiet environment.

Freshness

The dataset was last updated on 2025-04-24.

Geography

Speakers are from France, Canada, and Africa.