How would one exactly deploy a working audio classification model from Edge Impulse for mobile phone or web app applications (not the edge impulse mobile deployment kit). Assuming both on-device and backend scenarios? Are the preprocessing blocks built it and can i just stream or send WAV files from the frontend for inference?
Hi @flautistic
Web socket I believe is where you should look for where to send the samples - Remote management Protocol (WebSocket) | Edge Impulse API and SDKs Documentation
For running locally see -
Best
Eoin