Add option to use I2S slave audio rather than microphones as input to the ASR FFVA build config

lucianomartin commented 2 months ago

This PR contains the following changes:

it allows building a ffva_int_cyberon config where the audio fed into the intent engine is received via the I2S interface, rather than from the microphones.
the UART output of the FFVA Cyberon configs have been extended to include more information about the detections.

To build the FFVA Cyberon config with I2S input follow the steps below:

set appconfI2S_AUDIO_SAMPLE_RATE to either 16000 or 48000 Hz
set appconfUSE_I2S_INPUT to 1, this replaces the microphone audio with the I2S audio as input to the intent engine
build and run the build config example_ffva_int_cyberon_empty, no other pipelines have been tested so far

lucianomartin commented 1 month ago

Changes replaced by #361.

mbanth commented 1 month ago

@lucianomartin, why have you requested a review from me on a closed PR?

lucianomartin commented 1 month ago

@lucianomartin, why have you requested a review from me on a closed PR?

Sorry, it was a mistake. I meant to request a review for #361.

xmos / sln_voice