JarCasting 
Website | Documentation | EULA
Use the SpeechKit library to integrate speech recognition, text-to-speech, music identification, and Yandex voice activation into your Android mobile app. Voice control makes your app more user-friendly, especially for people who use it on the go or whose hands are busy. Without touching the screen, users can activate the desired feature with a single phrase.
SpeechKit supports the following languages for speech recognition and text-to-speech:
- Russian
- English
- Ukrainian
- Turkish
If the language you need isn't supported, SpeechKit allows you to switch to the Google speech recognition system.
If the number of voice commands accessing your application does not exceed 10,000 per day, you can use the basic version of the SpeechKit Mobile SDK for free. If you have more requests, you can either pay for the amount over the limit, or switch to the commercial pricing plan. The commercial plan removes the restrictions and adds extra functionality. For instance, you can create unique voices and speech models for specific contexts.
Where to start
- Find out what SpeechKit is, what functionality is available in the library, and how it can improve your application.
- Download SpeechKit samples and try to run them. This is a good way to quickly learn how to use the main components of the library.
- Read the Quick Start, which describes the configuration required for the library to work, along with recommended steps for fast and successful integration of SpeechKit in your application.
- Review the Documentation for the library's API: the main classes and their functions, features, and restrictions.
- If you have previously used SpeechKit, we recommend that you also read the Guide for migrating to SpeechKit 3.12.2, which explains what has changed since the latest available version.
Before you get started
Before you start working with APIs of Yandex services, you must create an API key:
- Go to the Developer Dashboard.
- Click
Get key. - Enter a name for the key and select
SpeechKit Mobile SDKin the list. - Enter your information and the type of project.
- Click
Submit.
After the API key is generated, you will be redirected to the Developer Dashboard. Under Your API keys, you can see all the keys that you have created.
Installation
You can use Maven to add SpeechKit to a project. To do this, open the build.gradle file in the project. Add the Maven Central repository to the list of available repositories. In the repositories section, type:
repositories {
...
mavenCentral()
}
Open the build.gradle file for the application (module). In the dependencies section, add the dependency:
dependencies {
...
compile 'com.yandex.android:speechkit:3.12.2'
}
In the build.gradle file for the application (module), go to the defaultConfig section and enter the minimum Android API version required by your application (14 or later):
defaultConfig {
...
minSdkVersion 14
}
Requirements
Minimum system requirements:
- SDK version 14
- Android Studio 2.3.3 +
Architecture
Global settings
SpeechKitLocationProviderEventLoggerLogLevel
Speech recognition
OnlineRecognizerOnlineRecognizer.BuilderLanguageOnlineModelSoundFormat
RecognizerRecognizerListenerTrackArtist
RecognitionRecognitionHypothesisRecognitionWord
BiometryBiometryGroupBiometryEmotionLanguageScore
RecognizerActivityAudioProcessingModeRecognizerUIEarconsRecognizerUIEarcons.BuilderDefaultEarconsBundle
Speech synthesis
OnlineVocalizerOnlineVocalizer.BuilderLanguageVoiceEmotionSoundFormatSoundQuality
VocalizerTextSynthesizingMode
VocalizerListenerSynthesisSoundBuffer
Voice activation
PhraseSpotterPhraseSpotter.BuilderPhraseSpotterListener
Additional functionality
AudioSourceAutoStartStopAudioSourceManualStartStopAudioSourceAudioSourceListener
UniProxySession
Usage
SpeechKit
Singleton class for configuring and controlling the library. You don't need to explicitly create or destroy instances of the SpeechKit class. To access an object, use the getInstance method, which creates an instance of the SpeechKit class when accessed the first time. This instance is destroyed when the application closes.
Before using any of the SpeechKit functionality, you must configure SpeechKit using the API key (for more information, see Before you get started).
SpeechKit.getInstance().init(getApplicationContext(), "developer_api_key")
OnlineRecognizer
OnlineRecognizer recognizer = new OnlineRecognizer.Builder(Language.ENGLISH, OnlineModel.QUERIES, this)
.setDisableAntimat(false)
.setEnablePunctuation(true)
.build();
recognizer.prepare();
recognizer.startRecording();
public void onPartialResults(@NonNull Recognizer recognizer, @NonNull Recognition recognition, boolean endOfUtterance) {
System.out.println("Partial result: " + results);
if (endOfUtterance) {
System.out.println("Partial result: " + recognition.getRequestId());
}
}
OnlineVocalizer
OnlineVocalizer.Builder vocalizer = new OnlineVocalizer.Builder(Language.ENGLISH, this)
.setEmotion(Emotion.GOOD)
.setVoice(Voice.ERMIL)
.build();
vocalizer.prepare();
vocalizer.synthesize("What's up kid?", Vocalizer.TextSynthesizingMode.APPEND);
Credits
The SpeechKit Mobile SDK is created and developed by Yandex.
License
The licensing terms for using SpeechKit are described in the Terms of Use. If you do not agree to any of the terms described in the license agreement, you cannot use SpeechKit.
Contacts
- If you have any general questions about the library, look for answers in the FAQ.
- If you found a bug in the library, and you can provide the exact steps for reproducing it or logs containing the necessary data, write to us via the feedback form.
- If you need help, contact us at [email protected].
- If you need to increase the maximum number of requests for speech recognition and/or speech synthesis, send an inquiry to [email protected].
