Hi All
We are working on developing an open source Voice Assistant that is based on the XU316 and the sln_voice FFVA example
I'm wondering how we can go about integrating wake word detection into the FFVA example https://github.com/xmos/sln_voice/tree/ ... mples/ffva
Initially any wake word will be sufficient for testing, but it would be nice to add custom wake word models. Are there any open source model trainers similar to Snowboy that would work with this system?
Kind regards
Alextrical
			
			
									
							
		XU316 wakeword in FFVA Topic is solved
- 
				Alextrical
														 - Member
 - Posts: 10
 - Joined: Sun Mar 24, 2024 9:47 pm
 
- 
				upav
														Verified - Active Member
 - Posts: 36
 - Joined: Wed May 22, 2024 3:30 pm
 
Hey Alextrical,
In the most recent version of the library (2.3.0) we've added a Cyberon integration for the int variant of FFVA. I don't think it made it to the website yet but you can see it on github.
Cyberon have their own tool for generating model datasets, which I think is what you are looking for.
Cheers,
Pavel
			
			
									
										In the most recent version of the library (2.3.0) we've added a Cyberon integration for the int variant of FFVA. I don't think it made it to the website yet but you can see it on github.
Cyberon have their own tool for generating model datasets, which I think is what you are looking for.
Cheers,
Pavel
Pavel
xmos software engineer
						xmos software engineer
