Simple extension of microWakeWord that uploads a few seconds of audio before each wake word detection. These are false positives if you never purposefully speak the wake word.
- Install the companion add-on
- Modify the
voicekit_fpc.yamlfile and set<HA IP ADDRESS>to the IP address of your Home Assistant machine - Flash the ESPHome YAML to the VoiceKit
Every time you speak the wake word ("ok nabu"), the VoiceKit should turn on the LEDs and upload audio to the server.
In Home Assistant, check the /share/voicekit-fpc directory for WAVE files.
Instead of using the HA add-on, you can just run server.py anywhere that has Python available. WAVE files will be saved in the current directory. See python3 server.py --help for more options.