Automatic Phoneme Recognition for Disordered Speech

The PSST Challenge dataset focuses on high-accuracy automatic phoneme recognition of disordered speech in a diagnostic context. It was created by jimregan and last updated in September 2024.

Use Cases

• Train models for automatic phoneme recognition on disordered speech samples.
• Develop diagnostic tools for speech and language disorders using phoneme-level analysis.
• Benchmark phoneme recognition accuracy for clinical applications in speech pathology.

Strengths

• Dataset is focused on a technically-challenging and clinically important task.
• Last update was recent, on September 10, 2024.

Limitations

• No information is provided on the dataset's size, number of rows, or specific features.
• The specific data format, columns, and sample data are unavailable for assessment.

Provenance

Source

huggingface

Freshness

Last updated 2024-09-10.