Datasets:
Tasks:
Automatic Speech Recognition
Formats:
soundfolder
Languages:
Chinese
Size:
10K - 100K
License:
metadata
license: cc-by-4.0
task_categories:
- automatic-speech-recognition
language:
- zh
pretty_name: ATC Generated Mandarin and Shaanxi Radio Speech
size_categories:
- 10K<n<100K
ATC Generated
Synthetic aviation-command speech generated with CosyVoice.
Contents
- 10,000 clean wav files.
- 10,000 10dB radio/noisy wav files.
- 5 reference speakers.
- Accents: standard Mandarin and Shaanxi dialect instruction.
Metadata
metadata.csv contains file_name, audio, utt_id, text, tts_text, speaker_id, accent, snr, noise_type, radio_effect, and related fields. The text column keeps Arabic numerals for ASR labels; tts_text stores the spoken-form text used for TTS.