Hello…
I have trained a model on bunch of sentences that were 2 seconds long…
like How’s the weather out there.
The window size was kept 1 second for the training and shift was 0.5 seconds.
So the arduino microphone (non continuous) example records for 1 second and make decision.
What if i want the arduino to listen for 2 seconds (sample size) and make decision. Would i have to train the model with window size = 2?