Might be interesting to train the model using user’s own voice. Maybe this would help filtering out co-workers in open office or family members. Maybe you coul

Might be interesting to train the model using user’s own voice. Maybe this would help filtering out co-workers in open office or family members. Maybe you could also use this personal model to hide very short network interruptions. Other party could use this model to constantly predict my next piece of audio and switch to prediction in case packet is lost.