How to implement Android TTS via MS Java SDK package (android) - Speech to Text

最新推荐文章于 2024-04-10 12:19:45 发布

rib-pet

最新推荐文章于 2024-04-10 12:19:45 发布

阅读量513

点赞数

CC 4.0 BY-SA版权

分类专栏： Android Java Azure Cognitive Services

tony-nf.peng

本文链接：https://blog.youkuaiyun.com/ningfeng8899/article/details/86765026

Java 同时被 3 个专栏收录

21 篇文章

订阅专栏

Android

5 篇文章

订阅专栏

Azure Cognitive Services

2 篇文章

订阅专栏

本文介绍如何在Android应用中使用Microsoft Azure的认知服务SDK进行语音到文本的转换。通过设置正确的订阅密钥和服务区域，可以调整语言和文化，实现不同语言的语音识别。文章还提供了如何开始和停止连续识别，以及如何在识别过程中获取实时结果的示例代码。

Speech to Text package:

package com.microsoft.cognitiveservices.speech;

Add following line in your [build.gradle - app]

implementation 'com.microsoft.cognitiveservices.speech:client-sdk:1.1.0'

in you specified project, you can add following code:

 config = SpeechConfig.fromSubscription(speechSubscriptionKey, serviceRegion);
            assert (config != null);
            // adjust different language culture
            clientContext = JSON.parseObject(getSharedPreference(ConstString.CLIENT_CONTEXT, 
            config.setSpeechRecognitionLanguage("zh-CN");
            config.setOutputFormat(OutputFormat.Detailed);
            recognizer = new SpeechRecognizer(config);

Note that You must set proper recognition language before you created "SpeechRecognizer" instance. Otherwise, the language will not be changed. For speechSubScriptionKey, and service Region, please get it from Azure with your azure account. (SpeechRecognizer has default language "en-US")

If you are going to retrieve one time result. You can call following result.

recognizer.recognizeOnceAsync()

What if you want to get result while starting recognition, you can call

recognizer.startContinuousRecognitionAsync();

and apply event "recognizing" to retrieve runtime result.

recognizer.recognizing.addEventListener(new EventHandler<SpeechRecognitionEventArgs>() {
                @Override
                public void onEvent(Object o, SpeechRecognitionEventArgs speechRecognitionEventArgs) {
                    text = speechRecognitionEventArgs.getResult().getText();
                    Log.d("recognizer.recognizing:", text);

                
                    setText(text);
                }
            });

In the end, Please do not forget to call close method of recognizer object in your onDestory method of Activity.

   @Override
    protected void onDestroy() {

        // in the end, recognizer need to close
        if (recognizer != null)
            recognizer.close();

        super.onDestroy();
    }

Btw: https://github.com/Azure-Samples/cognitive-services-speech-sdk/tree/master/samples/java/android/sdkdemo

Please click like, if you feel it's working on your project.