Speech Recognition using Google Cloud Speech API Build Environment: Android Studio 3.1.3 Android Platform SDK 26 Gradle Build 4.4 Android SDK build Tools 27.0.3 Java JDK 1.8
Dependencies:
dependencies { //Google Cloud Speech implementation 'com.google.api-client:google-api-client-android:1.22.0' implementation 'com.google.apis:google-api-services-speech:v1beta1-rev336-1.22.0' implementation 'com.google.apis:google-api-services-language:v1beta2-rev6-1.22.0' implementation 'com.google.code.findbugs:jsr305:2.0.1'
implementation project(':commons-io-2.6')//IO File Library
implementation 'info.hoang8f:fbutton:1.0.5'// Button design
implementation 'com.android.support:design:26.1.0'//UI design
implementation 'com.jaredrummler:material-spinner:1.2.5'//Spinner for language selection
}
Compile options:
compileOptions { sourceCompatibility JavaVersion.VERSION_1_8 //JDK 1.8 targetCompatibility JavaVersion.VERSION_1_8 //JDK 1.8 }
Gradle Configuration:
compileSdkVersion 26 defaultConfig { applicationId "com.example.sk.exp" minSdkVersion 23 targetSdkVersion 26 versionCode 1 versionName "1.0" testInstrumentationRunner "android.support.test.runner.AndroidJUnitRunner"
Material spinner Library:
Info: https://github.com/jaredrummler/MaterialSpinner Import: app:build:gradle -> dependencies
implementation 'com.jaredrummler:material-spinner:1.2.5'
Importing Commons IO Library:
Download Link: https://commons.apache.org/proper/commons-io/ Import: Extract the binary -> locate commons-io-2.6.jar Add dependency in app:build:gradle : implementation project(':commons-io-2.6')
Code:
Declare the API KEY to a String in MainActivity.java private final String CLOUD_API_KEY = "API KEY HERE";
The AudRec.java class generates audio.wav for the recording and replaces this file if new audio is recorded, this behaviour can be changed in getFilename() in AudRec.java.
The File Format used is wav with PCM encoding.
Other formats can be used with encoding specifications as mentioned in the link below: https://cloud.google.com/speech-to-text/docs/encoding