The previous article describes the basic concepts of speech recognition in Kinect, as well as some of the terminology used in speech processing. Examples of audio recordings using the Kinect microphone array illustrate the core object and configuration of the audio processing in Kinect. This article will continue to introduce speech recognition in Kinect, with two small examples to show direction recognition and voice command recognition in speech recognition.
1. Using a directional microphone for velocity tracking (Beam tracking for a directional microphone)
You can use these 4 microphones to simulate the effect of a directional microphone, called beam tracing (beam tracking), for which we create a new WPF project, as follows:
1. Create a WPF project named Kinectfindaudiodirection.
2. Add references to Microsoft.Kinect.dll and Microsoft.Speech.dll.
3. Change the name of the main form to "Find Audio Direction"
4. Draw a vertical, slender rectangle in the main form.
The slender rectangle on the interface is used to indicate the voice direction of the speaker detected at a given moment. The rectangle has a rotational transformation that swings around the vertical axis to indicate the different source direction of the sound. Front page code:
<rectangle fill= "#1BA78B" horizontalalignment= "left" margin= "240,41,0,39" stroke= "Black" width= "10" rendertransformorigin= "0.5,0" >
<Rectangle.RenderTransform>
<TransformGroup>
< scaletransform/>
<SkewTransform/>
<rotatetransform angle= "{Binding beamangle}"/>
< translatetransform/>
</TransformGroup>
</Rectangle.RenderTransform>
</rectangle >