Commit 5d4ac31913e2 ("audio_recorder: allow setting wake word") introduced the ability to select which wake word to use at runtime. This feature was dropped in a later commit. Restore it.
Fixes: 3f6abbb15b85 ("audio_recorder: support new feature of esp-sr (break change of audio_recorder)")
We use this feature in Willow to allow the user to select which wake word to use without having to reflash the model partition. Responding to "Hi ESP" when the user configured Alexa is confusing, so we'd like to have this feature restored. Having multiple wake words enabled also increases the chance of false positive wakes, at least theoretically.
Commit 5d4ac31913e2 ("audio_recorder: allow setting wake word") introduced the ability to select which wake word to use at runtime. This feature was dropped in a later commit. Restore it.
Fixes: 3f6abbb15b85 ("audio_recorder: support new feature of esp-sr (break change of
audio_recorder
)")We use this feature in Willow to allow the user to select which wake word to use without having to reflash the model partition. Responding to "Hi ESP" when the user configured Alexa is confusing, so we'd like to have this feature restored. Having multiple wake words enabled also increases the chance of false positive wakes, at least theoretically.