Speaker Identification on 8Khz voice
#1
by
giusarno
- opened
I read that model trained with data at 16Khz will not work well on data sampled at 8Khz. Is it the case for this model ?
How to use this model with google cloud functions using python?
I have also encountered this problem and found that the performance of 16kHz audio files is better. I am now using 8kHz audio files to resample to 16kHz, and the effect is also relatively poor.
I don't know how to make it perform better when I only have 8kHz sound.