There are two systems families : speaker independent and dependent.
The EasyVR3 supports them both but for SI You must acquire an utility for more than 150$.
Now there are two other SI systems :
- SimpleVR3 by Elechouse (only US English and Chinese), rather cheap and reliable if You choose carefully Your vocabulary and don't have more than 4 or 5 expressions active simultaneously in each context ; unfortunately it has no adjustable "strictness level" (in English) and it needs that there is absolutely no background noise : despite of the fact that only one train is running at once on my layout, there is during the pauses the noise of the motors of the funfair which are started and stopped "asynchronously" by a dedicated ATTiny13 and this "destabilizes" the recognition.
- Another system, Audeme's MOVI, completely dedicated to Arduino (at least 7. 5V for power ! ) needs a wake-up word and seems rather slow (on YouTube).
- Forget the LD3320 and other similar systems, not well documented and completely "Chinese language contaminated" .
The EasyVR3 works well in SD mode, is rather unsensitive to background noise (in standard sensitivity mode) nut needs training the keywords (in my case, 7) at least twice to be safe.
Or does anybody know another SI system, but not online (i.e. Siri-like or Alexa-like) : I don't wanna interface (if even possible ! ) with the net ... I am afraid that training keywords would be felt as "has been" by kids who use the online systems !
Lionel, too, sells such a system (Look on YouTube for "Lionel Voice Control" ) but it runs either on a tablet or a smartphone, so I think that it uses online recognition.
Last edit: Bertrand BAROTH 2022-09-05
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
There are two systems families : speaker independent and dependent.
The EasyVR3 supports them both but for SI You must acquire an utility for more than 150$.
Now there are two other SI systems :
- SimpleVR3 by Elechouse (only US English and Chinese), rather cheap and reliable if You choose carefully Your vocabulary and don't have more than 4 or 5 expressions active simultaneously in each context ; unfortunately it has no adjustable "strictness level" (in English) and it needs that there is absolutely no background noise : despite of the fact that only one train is running at once on my layout, there is during the pauses the noise of the motors of the funfair which are started and stopped "asynchronously" by a dedicated ATTiny13 and this "destabilizes" the recognition.
- Another system, Audeme's MOVI, completely dedicated to Arduino (at least 7. 5V for power ! ) needs a wake-up word and seems rather slow (on YouTube).
- Forget the LD3320 and other similar systems, not well documented and completely "Chinese language contaminated" .
The EasyVR3 works well in SD mode, is rather unsensitive to background noise (in standard sensitivity mode) nut needs training the keywords (in my case, 7) at least twice to be safe.
Or does anybody know another SI system, but not online (i.e. Siri-like or Alexa-like) : I don't wanna interface (if even possible ! ) with the net ... I am afraid that training keywords would be felt as "has been" by kids who use the online systems !
Lionel, too, sells such a system (Look on YouTube for "Lionel Voice Control" ) but it runs either on a tablet or a smartphone, so I think that it uses online recognition.
Last edit: Bertrand BAROTH 2022-09-05