2

I am using Windows IOT (Windows Universal App) on a dragonboard along with Windows.Media.SpeechSynthesis and Windows.Media.SpeechRecognition.

The problem I am facing is that the Speech Synthesis is picked up by the SpeechRecognition of the system.

I do not want to disable the recognition while the text to speech is running since I have to keep the system in an 'always listening state', so I can accept commands even while it is talking.

Is there any Windows IOT class etc. that would prevent such a conflict? Or are there any other solutions that I could employ?

Thank you.

Richard Chambers
  • 16,643
  • 4
  • 81
  • 106

1 Answers1

0

I create programs in winforms c#. When I use System.Speech, I always define System.Speech.Recognition and System.Speech.Synthesis. I have not tried it on a universal side. However, the only problem I have come to use example, Alexis.SpeakAsync("....."); Alexis is the name of one of my programs. I have never had any conflict between Recognition and Synthesis cross over.. This is an idea you can try if you have the System.Speech reference.

(add) you can always add a small piece of code I always use that mutes the microphone after so many seconds and you say a keyword for it to listen to commands.. This helpful when your talking to some one, the system will listen for the keyword then listen for commands.

ivansystems
  • 117
  • 12
  • That is the problem exactly. For example, currently the system starts off by saying, please say 'Wake up' to activate me. While it says this, the mic picks up the words wake up and automatically activates. I understand what you meant by mute the mic while it is speaking, but I cannot do that since I need the system to always be on a listening state, to issue commands like 'stop talking' etc. –  Feb 11 '16 at 15:48
  • Well let me ask. Why did you go with System.Media instead of System.Speech? – ivansystems Feb 11 '16 at 16:00
  • Since I am using it on Windows IOT on the dragon board, the System files are not available since WIndows.Media, I believe has a lighter footprint. But I am pretty sure the source is the same. –  Feb 11 '16 at 16:07
  • I really want to show you my system, because it is a good example, I have a video on my website of me and my system having a conversation. I am thinking you are wanting the same thing. Look at the videos at the bottom of the home page.. Let me know if that's what your going for. www.ivansystemsinc.ml if that's what your going for then I can give yo an example code in c# winforms you will just have to convert it – ivansystems Feb 11 '16 at 16:12
  • Your website is great. It is exactly the kind of stuff I want to work with. Do you have a email ID I could personally contact you at? –  Feb 11 '16 at 16:19
  • ivansystems@outlook.com also another trick is, if your using built in mics on your system it will cause a lot of feedback, if you use a headset then it eliminates all the extra feedback... Watch the first video and you will see it's actually quite accurate the way I built the system. – ivansystems Feb 11 '16 at 16:22
  • I am not using system mics. I have a separate mic and a separate speaker. Is there no software fix available that would solve this conflict problem? –  Feb 11 '16 at 16:34
  • I sent you an email, but with software fixes, this has to do more with the coding side. You will just have to tweak the code. I sent an email to you that is more in-depth – ivansystems Feb 11 '16 at 17:05