20 hours of audio recordings capturing communication between air traffic controllers and pilots, provided by the University of West Bohemia. The data includes manual transcriptions and speaker labels identifying the role of the speaker as either pilot or controller.
Use Cases
- Train domain-specific automatic speech recognition (ASR) models using the audio recordings and manual transcriptions
- Build speaker classification models to identify roles using the pilot/controller speaker labels
- Analyze aviation communication patterns and phraseology through the provided manual transcriptions
Strengths
- 20 hours of audio recordings of pilot-controller communications
- Manual transcriptions for all recorded speech segments
- Speaker labels identifying roles as either 'pilot' or 'controller'
- Sourced from the University of West Bohemia, Department of Cybernetics