gabrielmittag / NISQA

NISQA - Non-Intrusive Speech Quality and TTS Naturalness Assessment
MIT License
663 stars 117 forks source link

feat: Add Descriptive Statistics and Charts #37

Open Hadlay-Zhang opened 1 year ago

Hadlay-Zhang commented 1 year ago

Hi Gabriel,

Thanks a lot for your work on the NISQA. I have been using the Speech Quality Prediction in my work related to testing audio quality of distinct devices, and it has been incredibly helpful. Here are some new features I have made to prediction:

Added descriptive tools for Speech Quality Prediction:

This change helps provide better assistance in Speech Quality Prediction, especially for repeated averaging of results of the same test sample in practical Audio Testing Scenarios. Also, output name of CSV is added just like #30.