Hi all,
I'm the person behind openwakeword.com and microwakeword.com (the websites, not the GitHub libraries). Over time the free public wake word database has become fairly complete, and I think it would be great to find a way to make it available outside the websites too.
I'd love ideas on how a wake word database could best fit into HA for frictionless wake word selection.
Keep in mind that:
-
The database is not curated
-
Accents can affect detection accuracy
So it would help to have a test interface somewhere, like the one I built into the website's wake word library.
Curious to hear your ideas!
Stefano