Main Site Documentation

Speech Recognition on NETMF


For those how are interested in, have a look on this :

I don’t know if we already posted about this topic, so I apologize if his post is already on the forum .



I don’t think this is what you think it is…

It looks like you take a string and send it to a Microsoft service, who then converts it to a MP3 byte array, which you can then play with a music module.


@ Mike -

yeee, this looks like byte -> speak. I am looking for speak -> byte ;D


Hrrrggg !

speak to byte will be more difficult on NETMF hardware, as it requires memory and signal processing !


Unless if you record audio and send it to data center to process it. That’s how Siri and alike services work.


@ iamin -

Yes, but however it will required a piece of memory that is not available on every modules !


I guess you could try to do something like this: record 1 second save it to memory card, record another second - save it and so on.


I think the only way to achieve this is have a opposite web service as in the subject where you can post the audio fragment and receive the result back.