Is it in any way possible to automatically split a video track based on it's audio clip peaks. 

Like for example if i have audio commentary for a video and i wanna cut out and delete all the footage where i don't say anything, I would go manually through the video and cut out the pieces where i can see that there is no commentary audio playing based on the waveform.

Is there some automatic way to detect where the waveform is ~0 volume, versus ~20 volume and up which would indicate commentary, and also to specify some buffer space between the cuts to leave a little breathing room between pieces of commentary?

or is there no way to do something this sophisticated?