Continuous speech recognition in android app without google popup - java

I am working on a project on "Home Automation" system with Android app and Microcontroller , connecting them through Bluetooth module. I have incorporated "speech to text" for voice commands. Although , it works well with the built-in google speech recognition api. All I need is a continuous speech recognition and without the google voice pop-up.
I have seen a similar app on playstore "speechnotes" that works exactly i wish to develop. And it works totally offline.

You should try Droid Speech it supports continuous voice recognition and also will work offline if the speech package for the required language is installed in the device.
Setting the library in your project is very simple and with a few lines of code you can easily start using speech recognition.
An example,
Add the below in your Gradle file,
compile 'com.github.vikramezhil:DroidSpeech:v2.0.3’
In your Activity,
DroidSpeech droidSpeech = new DroidSpeech(this, null);
droidSpeech.setOnDroidSpeechListener(this);
To start droid speech to listen to user voice call the method,
droidSpeech.startDroidSpeechRecognition();
The speech result will be triggered at,
#Override
public void onDroidSpeechFinalResult(String finalSpeechResult)
{
// Do whatever you want with the speech result
}

Related

Is there a way to open an android mobile app using voice command without the need of publishing the app in play store?

I am trying to launch the android app using voice command without the need of Google's speech to text. Is there any other way to launch the app, since I am developing this app for visually impaired users? They need to open the app in the most easiest possible way. Thanks in advance for your feedback ;)
I tried opening the app using Google speech to text, but other apps are being recognized. Then I found out that the app needs to be published in the play store in order to have its own voice command.

Send text strings from Android application/plugin to Unity application

I need to add non-Google offline voice control capability to a Unity3D project. I have started with snips.ai and successfully built an Android app following their tutorial and modified it to output the understood intent from snips to a text box on screen. What I really need is to get this text string to send to my Unity application.
I think I should be building a native Android Java plugin for Unity which will start at runtime and listen for voice commands. Every time it gets a listen event I need it to send a string of the captured intent to the Unity project which is always waiting for these strings.
I know I can use something like this https://gist.github.com/markcastle/a1530efd5278607862fec09df89e59db to make a static call to a native plugin, but I want to basically have Unity listening constantly to the plugin for strings which it can interpret.
Any help, as always, is much appreciated!

WP10 Detected Bluetooth device functions

As all users of Windows Phone 10 know, the Bluetooth connectivity on this system is cutted to the simplest for end-user, which ends for programmers as a horror to create anything. When connecting to a device, system automatically looks for a "functions" that device has, for ex. audio. The questions are as following:1. How to define a function like that on external Android device? The main goal is, to control all types of music playback from it (ex. Groove music playing)2. Where can i find a list of things like this, for future app functions?
And to get ahead of answers, I already tried the easier ways, which came to nothing due to UWP limits of controlling other apps playback.
Main target of project is to make a xamarin c# (or java, I can "translate" from one to another with ease) android app for a smartwatch and in c# uwp for windows phone to do most common tasks, like - as mentioned before - music playback control and notifications reciever.
Based on your description, you want to develop an Android app to send Bluetooth command to control the music player of a windows phone device, then it depends on your Bluetooth of your android device.
But control music player by Bluetooth need the profile Avrcp I have check the google Bluetooth API I did not find the profile of Avrcp, I think it is may be hard to send the Avrcp command at the android application level.
And Android is source opened, many device factories prefer to customize their own Android system, it is possible that the device uses a Bluetooth module as a serial device, then the official Bluetooth APIs are not suitable for this scenario, the develop protocol of this Bluetooth module is in need.

Speech Recognizer on HTC One M7

I wrote a speech recognition app using android's built-in speech recognition classes. The following exception shows up in my developer console when the startListening function is called on the speech recognizer object obtained using createSpeechRecognizer(context) function.
SecurityException:
java.lang.SecurityException: Not allowed to bind to service Intent
{ act=android.speech.RecognitionService
cmp=com.htc.android.voicedictation/.VoiceDictationService
}
Any ideas why this is happening and how to fix this?
Tested working on Nexus 7 & HTC ONE X.
Failed on HTC ONE M7 (4.2.2)
Found the root cause of this issue from this link: https://groups.google.com/forum/#!topic/joaomgcd/qcSfL2-SIfk
Solution: Google voice typing must be enabled on a device.
Steps to enable Google voice typing on HTC One device:
http://eguides.sprint.com/support/eguides/htconem8/index.html#htc_one_m8_ug/keyboard_and_input_methods_google_voice_typing.html
I have faced the same issue. I solved it by Going to Settings-> Language & keyboard -> Voice Input-> Then enable the 'Basic Google recognition', as this radio button was disabled in my same HTC phone.

SpeechRecognition API: How to get the voice features (MEL Coefficients)

I was going to implement a speaker verification app for Android, and was wondering if there would be a way to get the voice features (MEL Coefficients) from the Android's Speech Recognition module?
Please note that speaker verification is slightly different than speech recognition, so that's why I'm not using the speech recognition directly.
No, you can not compute MFCC from audio stream with Android API. You can use external library for that, for example CMUSphinx
You can also try existing Java-only libraries for voice verification for Android like
Recognito

Categories