Repeated wake word activation.

[mushie4282]

I disabled "hey mycroft" hotwords and implemented my own. After rebooting, OVOS continuously activates and transcribed random words. Does anyone know the diagnosis for this issue?

Here is my GitHub that links my trained wake word models and user config files used.

• https://github.com/SeniorDesign-RC-LB/wakeWordModels/tree/main?tab=readme-ov-file

I am running Raspberry Pi 4B 2GB with RaspOVOS img (headless).

[builderjer]

Are you using openwakeword?

There is a threshold value you can play with.

Check the readme for more details

[mushie4282]

Yes, I'm using openWakeWord. After changing the threshold, I'm still coming across the same problem. I've attached my issue from viewing the voice log.

[JarbasAl]

it seems like the trained models just are not very good

[mushie4282]

@JarbasAl, probably.
I wanted to post this discussion to see if anyone else is running into the same issue. Does anyone have tips or advice on training wake word models on Window machines?

[mushie4282]

Update
I swapped my model to a pretrained model (android.tflite) to use with ovos-ww-plugin-precise-lite. https://github.com/OpenVoiceOS/precise-lite-models/tree/master/wakewords/en
I'm still having the same issue, where continuous wake word activation occurs.

[builderjer]

It sounds like the model was trained with synthetic voices. You might try changing some settings