Speech recognition [win10]

advanced
#1

Hi,

I was wondering if anyone’s used Microsoft’s speech recognition services.
https://docs.microsoft.com/en-us/windows/uwp/input-and-devices/speech-recognition

I’d be curious if there’s a possibility to set-up this up with openFrameworks. I can run a c++ example from Microsoft and it works just fine, but I have too little knowledge of C++ and Visual studio (2015) to try and port it to OF. I’ve tried using https://msopentech.com/blog/2014/07/03/openframeworks-developers-can-now-create-universal-apps-for-windows/
for setting up an OF project with it, but after a few steps into the process I get a bunch of errors and that’s where things stop.

As usual, any insights are welcome.

Thanks!

#2

I was trying to get this to work myself but I was getting errors. Good luck man.

#3

following a tutorial here -> https://blogs.windows.com/buildingapps/2017/01/25/calling-windows-10-apis-desktop-application/#78vQMtCLjB5ZWQHc.97 I’m able to add namespaces needed. Unfortunately that’s it :frowning:

e.g. adding:

	Windows::UI::Core::CoreDispatcher^ dispatcher;
	Windows::Media::SpeechRecognition::SpeechRecognizer^ speechRecognizer;
	std::wstringstream dictatedTextBuilder;

to ofApp.h works fine, but as soon as I try to create any instances I get into trouble. E.g.
speechRecognizer = ref new SpeechRecognizer(SpeechRecognizer::SystemSpeechLanguage);
just builds fine, but causes the console to display “press any key to continue. . .” message after running.
No OF window to be seen. Any ideas? Ways to debug?

#4

I’m wondering if these things have to be run in a main UI thread… Not sure if I’m just making shit up :slight_smile:

I can run this code in main.cpp

Windows::UI::Core::CoreDispatcher^ dispatcher = CoreWindow::GetForCurrentThread()->Dispatcher;
Windows::Media::SpeechRecognition::SpeechRecognizer^ speechRecognizer = ref new Windows::Media::SpeechRecognition::SpeechRecognizer();