AKA
AKA

Reputation: 6935

Is there any ways to use own speech recognition engine with android.speech module?

I have implemented a demo speech recognition service my own. I deployed it on a server. Now I can access the service by using an API (http://192.168.1.100/ASR/demoSpeechToText). I am passing audio file in wav format and language parameter to the API. The API is working fine.

I wanted to integrate this with an Android application. I implemented it as a simple API call from the app which will be called after recording a voice sample.

I wanted to explore further to check the possibility of integrating it with the android.speech module (https://developer.android.com/reference/android/speech/package-summary.html).

So my question is, Is there any way that I can extend some classes from audio.speech module and write my own class to call my speech recognition API than the default one? I have seen this one: https://developer.android.com/reference/android/speech/SpeechRecognizer.html, but don't know if possible to use with my own API.

Upvotes: 3

Views: 689

Answers (1)

Kaarel
Kaarel

Reputation: 10672

A very minimal extension of RecognitionService is e.g. https://github.com/Kaljurand/speechdemo1 (simply shows which classes need to be extended, does not actually do any recording/recognizing), a more complete one is https://github.com/Kaljurand/K6nele (records audio and streams it to an existing cloud service, but does not support wav-files).

Upvotes: 1

Related Questions