what algorithm is used in the audio labeler app?

Hi,

I've developed a labeler algorithm to ectract useful info from time series. I've tested my algorithm against the soeech detector feature in the audio labeler app.

Both algorithms (mine and MATLAB's) have different behaviours, MATLAB's catched some that I miss and the other way round. I would like to know how they built this feature to improve my code.

Could someone reference me to a publication or a reference page with this info?

Thanks,

Uri

Answer 1

[jibrahim]

Hi Uri,

The 'speech detector' automatic labeling is based on the function detectSpeech, which is documented in Audio Toolbox:

mathworks.com/help/audio/ref/detectspeech.html?s_tid=srchtitle

The reference on which the function is based can be found in the doc page.