Using VAD feature

Hi, I want to save audio clips when a person speaks in a noisy indoor environment. I would like to use VAD feature in the respeaker mic array v2.0 to perform this task. Basically I would like to know how to tune in the parameters in the mic array and how often should i check the VAD feature when performing the above task. I would also like to know whether there exists a better way to perfrom this task as well.

P.S. : if possible, I would like to know the algorithm or the paper used to implement VAD in Respeaker mic array V2.0

HI there,

For the tuning, please refer to <LINK_TEXT text=“ … .0/#tuning”></LINK_TEXT>

Sorry, we signed the NDA with xmos about the algorithm. thanks for understanding.