Fongbe Speech Recognition Audio and Text Dataset

A collection of audio waveforms and corresponding transcriptions for Fongbe speech recognition. The audio data is sampled at 16,000 Hz. The dataset was created by godwinh and was last updated in May 2022.

Use Cases

• Train an automatic speech recognition model on audio waveforms and text transcriptions.
• Analyze phonetic patterns in Fongbe using the transcribed text data.
• Build a text-to-speech system by learning from the paired audio and text samples.

Strengths

• Audio data is provided as waveforms at a 16,000 Hz sample rate.
• Includes paired text transcriptions for each audio sample.
• Dataset was updated in May 2022.

Limitations

• The total number of audio-text pairs (rows) is unknown.
• The specific columns and data structure are not documented.
• Geographic origin and speaker demographics for the audio are unspecified.

Provenance

Source

Original repository transformed by godwinh.

Collection Method

Transformed from an original repository to include waveform values directly in a CSV file.

Freshness

2022-05-30