Modify the listen_node.py to capture a recording of the audio picked up by KEN's mic.
An option to enable/disable the recording is desirable.
Minimally, recording all the sound into a WAV file from startup until shutdown would help.
Recording only the sound when KEN is "listening" would also be useful, to eliminate the sections of sound when KEN is speaking.
Recording a separate annotations file indicating the times within the recording when events happened (sound started, voice detected, sound ended, speech started, speech ended) might also help with the analysis.
Modify the listen_node.py to capture a recording of the audio picked up by KEN's mic.
An option to enable/disable the recording is desirable.