You need a cue that's both audio and video.
So like Fezstar suggests time it with a button press. So in the menu you could press play on a countdown, so say "3 - 2 - 1 - press"
Or
If your commentary is recorded on the video you could simply just clap a few times, then you can sync up the audio claps with the facecam.
How do you record you commentary?