Fix SNR calculate error, total_snrlevels, requirements.txt

microsoft / MS-SNSD

The Microsoft Scalable Noisy Speech Dataset (MS-SNSD) is a noisy speech dataset that can scale to arbitrary sizes depending on the number of speakers, noise types, and Speech to Noise Ratio (SNR) levels desired.

MIT License

468 stars 142 forks source link

Fix SNR calculate error, total_snrlevels, requirements.txt #22

Open drumpt opened 2 years ago

drumpt commented 2 years ago

I think np.sqrt in line 68 of audiolib.py should be removed in order to get correct snr.
total_snrlevels should be a integer parameter in np.linspace(snr_lower, snr_upper, total_snrlevels) (line 47 of noisyspeech_synthesizer.py)
I removed unnecessary standard library "os" and changed glob to glob2 to accurately download libraries in requirements.txt

ghost commented 2 years ago

All CLA requirements met.