Instructions for using VAD function in the evaluation sdk

Here is an example of the commonly used Android and iOS platforms.

1.Load vad resource

Put the vad resource package in the same directory as provision file, and configure vad field in the interface parameters of engine creation：

Vad field parameter description

Name	Type	Required	Description
vad	object	false	Voice activity detection
- enable	int	false	Default 0. 1, means that the engine created this time loads the voice activity detection function module. 0, means that the engine created this time does not load the voice activity detection function.
- res	string	false	Vad resource path
- sampleRate	int	false	Audio sample rate in Hz
- strip	int	false	Whether to filter the silent frames in the beginning and end, default value is 1, recommend 0

Sample code
Android platform: please refer to the sample code in Step 1 to create engine in the Android sdk documentation;
iOS platform: please refer to the sample code in Step 1 to create engine in the iOS sdk documentation;

2. Configuration parameters

Set the VAD function when making a request

Name	Type	Required	Description
param	object	true	Evaluation content
- vad	object	false	Sound detection function
- - vadEnable	int	false	Default 0 1: Enable VAD. 0: Disable the VAD.
- - refDuration	int	false	Default 0 Set how long will the vad be disbaled after recording starts (unit: seconds)
- - speechLowSeek	int	false	Silence sensitivity, 20ms per frame, set N means after N*20ms of silent audio, it would return 2(time to stop engine)

Sample code
Android platform: please refer to the sample code in Step 2 launch evaluation request in the Android sdk documentation;
iOS platform: Please refer to the sample code in Step 2 launch evaluation request in the iOS sdk documentation;

3. Process result

Add the processing of different states of vad_status in the callback function

@Override
   public void onVad(String s, final EvalResult evalResult) 
	{

       //Update main thread UI
       runOnUiThread(new Runnable() 
		{
           @Override
           public void run() 
			{
               jsonResultTextEditor.setText(evalResult.text());
           }
       });

       try 
		{
           JSONObject json = new JSONObject(evalResult.text());
           if (json.has("vad_status"))) 
			{
               int status = json.optInt("vad_status");
               final int sound_intensity = json.optInt("sound_intensity");
               if (status == 2) //Detect that the user has stopped speaking, call the aiengine.stop interface to stop sending audio data, 
			                    //and wait for the engine to feedback scoring result
				{
                   runOnWorkerThread(new Runnable() 
					{
                       public void run() 
						{
                           RetValue retstop = aiengine.stop();
                           runOnUiThread(new Runnable() 
							{
                               public void run() 
								{
                                   if (recordButton.getText().equals(getText(R.string.stop))) 
									{
                                       recordButton.setText(R.string.record);
                                   }
                               }
                           });
                       }
                   });
               }
           } 
			else 
			{
               if (recordButton.getText().equals(getText(R.string.stop))) 
				{
                   recordButton.setText(R.string.record);
               }
           }
       } catch (JSONException e) 
		{
           Log.d(TAG, "pare result error!");
       }

   }