All events the VAD can emit.
A voiced segment has ended — the STT model should be flushed.
Voiced speech has been detected — frames should now be forwarded to STT.
All events the VAD can emit.