Voice Biometrics - Android

VDK features two Voice Biometrics SDKs: vsdk-tssv and vsdk-idvoice.

Conﬁguration

Voice biometrics engines must be conﬁgured before the program starts. Here is a complete setup for the vsdk-tssv SDK:

JSON

{
    "version": "2.0",
    "tssv": {	// Contrary to other technologies,
        "biometrics": { // biometrics paths are relative to the program's working directory!
            "generated_models_path": "data/models",
            "background_model_TI": "data/text-independent-16kHz.ubm",
            "background_model_TD": "data/text-dependent-16kHz.ubm"
        }
    }
}

Starting the engine

JAVA

com.vivoka.vsdk.Vsdk.init(mContext, "config/main.json", vsdkSuccess -> {
    if (vsdkSuccess)
    {
        com.vivoka.tssv.Engine.getInstance().init(mContext, engineSuccess -> {
            if (engineSuccess)
            {
                // at this point the BiometricEngine has been correctly initialized
            }
        });
    }
});

Creating a model

Models contain enrollment data that recognition operations need.

JAVA

// To create a text independant model
Model model = com.vivoka.tssv.Engine.getInstance().makeModel("test_ti", ModelType.TEXT_INDEPENDANT);

// To create a text dependant model
Model model = com.vivoka.tssv.Engine.getInstance().makeModel("test_ti", ModelType.TEXT_DEPENDANT);

Checking users in the model

If a model was previously created with the same name it will be loaded. You can check enrolled users with:

JAVA

if (model.getUsers().size() == 0)
{
  // No enrolled users
}

Adding a user to the model

You can either add raw audio data directly. After adding all the data for a given user, compile the model to ﬁnalize the enrollment process:

JAVA

//model.addRecord(String name, Buffer buffer);
model.addRecord("victorien", buffer);

The more data you give the model, the better the result will be.
Prefer to register the data in the condition of the use case of the model.

The format is preferred to be 16Khz mono-channel.

Performing authentication or identication

Both are covered in the same chapter as it is very similar:

JAVA

// Identificator
Identificator identificator = com.vivoka.tssv.Engine.getInstance().makeIdentificator("ident", model,
    new IStatusReporter<IdentificatorResultType, IdentificatorEventCode, IdentificatorErrorCode>() {
        @Override
        public void dispatchResult(IdentificatorResultType authenticatorResultType, String s, JSONObject jsonObject, boolean b) {
        String user = jsonObject.optString("id");
        }

        @Override
        public void dispatchEvent(IdentificatorEventCode identificatorEventCode, String s, String s1, int i) {}

        @Override
        public void dispatchError(ErrorType errorType, IdentificatorErrorCode identificatorErrorCode, String s, String s1) {}
    });

// Authenticator
Authenticator authenticator = com.vivoka.tssv.Engine.getInstance().makeAuthenticator("ident", model,
    new IStatusReporter<AuthenticatorResultType, AuthenticatorEventCode, AuthenticatorErrorCode>() {
        @Override
        public void dispatchResult(AuthenticatorResultType authenticatorResultType, String s, JSONObject jsonObject, boolean b) {
        String user = jsonObject.optString("id");
        }

        @Override
        public void dispatchEvent(AuthenticatorEventCode authenticatorEventCode, String s, String s1, int i) {}

        @Override
        public void dispatchError(ErrorType errorType, AuthenticatorErrorCode authenticatorErrorCode, String s, String s1) {}
    });
authenticator.setUserToRecognize("victorien");

The only difference is that the authentication can only recognize user “victorien” and the identiﬁcation can recognize every user enrolled in the model.

Different SDKs will give you different results, for example vsdk-idvoice reports varying results as it analyzes the audio, while vsdk-tssv only sends you result if the engine thinks it is acceptable (depending of the conﬁdence level you set).

We recommend that you try it out the application in real situation to select your custom minimum score required to satisfy your need in false rejection and false acceptation. But by default you can just check if the score is above 0.