This micro_speech_pico implementation is using the modifications introduced in MICRO_SPEECH implementation for PICO-TFLMICRO and requires the Raspberry Pi Pico Arduino core, for all RP2040 boards by Earle F. Philhower, to be installed and used in Arduino IDE.
The main change compared to older implementations is that the tensorflow/lite/experimental/microfrontend
is not used
and instead the new approach, with a separate TFLM model for the audio preprocessing is used.
See the official Micro Speech Example for details.
The default implementation in command_responder.cpp
writes out the name of the recognized command
to the serial port and turns on/off the default LED for the recognized yes/no commands.
Real applications will want to take some custom action instead, and should implement their own versions of this function.