Page 1 of 1

keyword recognition on VocalFusion Stereo Dev Kit for Amazon AVS

Posted: Thu May 17, 2018 11:39 am
by lucami
Hi,
how would one get a custom keyword recognition on this device? The device supports keyword "Alexa" when Amazon demo program is ran. I'd like to achieve something similar, but without Amazon Alexa service for now. When a certain keyword is produced (either through TV, speakers or a person), the device should pick it up and update a counter or something similar. Where would be a starting point to program something like this?

Thanks

Re: keyword recognition on VocalFusion Stereo Dev Kit for Amazon AVS

Posted: Fri May 18, 2018 12:13 pm
by mon2
Hi. You should contact your local XMOS support FAE or manager for assistance. Or consider to open a support ticket through xmos.com.

Our understanding is that there are 2 basic choices, Sensory and Kitt.AI - better to ask XMOS for clarification.

https://www.digitaltrends.com/home/sens ... word-news/

Do post your results here to improve this user forum.

http://www.xcore.com/viewtopic.php?f=3& ... ory#p31327

Re: keyword recognition on VocalFusion Stereo Dev Kit for Amazon AVS

Posted: Mon May 28, 2018 8:49 am
by lucami
You were correct. This is the reponse from xmos support:

We do not currently support multi-keyword recognition on xCORE however you might be able to do this by connecting the VFST board to a Raspberry Pi running Kitt.AI Snowboy : https://github.com/Kitt-AI/snowboy.