Closed gamename closed 7 months ago
I think it is possible. The key is that we need to collect enough meow sound samples.
I think it is possible. The key is that we need to collect enough meow sound samples. @feizi Great! Is this something I could create on my own?
Now we do not open source training pipeline. If you want to train a model by yourself, I suggest you to search [keyword spotting] in github. You may find some useful project. And then you can use esp-dl to deploy model on ESP32.
I'm interested in trying to use a cat's meow sound as a "wake word". Is that possible?
Thank you, -T