Android Studio内置语音识别库入门指南：从零开始构建语音交互应用

1次阅读

没有评论

共计 3323 个字符，预计需要花费 9 分钟才能阅读完成。

语音识别（Speech Recognition）是将人类语音转换为文本或命令的技术。在移动应用中，它能实现以下场景：

语音输入替代键盘输入
语音控制应用功能
实时语音转文字记录
无障碍辅助功能

Android 平台通过 SpeechRecognizer 类提供原生语音识别支持，无需集成第三方 SDK 即可实现基础功能。

Android 的语音识别能力主要依赖两个核心类：

SpeechRecognizer：处理语音识别流程
RecognitionListener：接收识别结果回调

使用限制：

必须联网（依赖 Google 语音服务）
每次识别最长持续约 5 分钟
不支持完全离线模式

在 AndroidManifest.xml 中添加：

<uses-permission android:name="android.permission.RECORD_AUDIO" />
<uses-permission android:name="android.permission.INTERNET" />

在 Activity 中添加以下代码：

private fun checkAudioPermission() {
    if (ContextCompat.checkSelfPermission(
            this,
            Manifest.permission.RECORD_AUDIO
        ) != PackageManager.PERMISSION_GRANTED
    ) {
        ActivityCompat.requestPermissions(
            this,
            arrayOf(Manifest.permission.RECORD_AUDIO),
            REQUEST_RECORD_AUDIO
        )
    }
}

class SpeechActivity : AppCompatActivity(), RecognitionListener {
    private lateinit var speechRecognizer: SpeechRecognizer

    override fun onCreate(savedInstanceState: Bundle?) {super.onCreate(savedInstanceState)
        setContentView(R.layout.activity_speech)

        if (SpeechRecognizer.isRecognitionAvailable(this)) {speechRecognizer = SpeechRecognizer.createSpeechRecognizer(this)
            speechRecognizer.setRecognitionListener(this)
        } else {Toast.makeText(this, "设备不支持语音识别", Toast.LENGTH_SHORT).show()}
    }
}

fun startListening(view: View) {val intent = Intent(RecognizerIntent.ACTION_RECOGNIZE_SPEECH).apply {putExtra(RecognizerIntent.EXTRA_LANGUAGE_MODEL, RecognizerIntent.LANGUAGE_MODEL_FREE_FORM)
        putExtra(RecognizerIntent.EXTRA_PROMPT, "请开始说话")
    }
    speechRecognizer.startListening(intent)
}

override fun onResults(results: Bundle?) {val matches = results?.getStringArrayList(SpeechRecognizer.RESULTS_RECOGNITION)
    matches?.let {if (it.isNotEmpty()) {val recognizedText = it[0]
            textViewResult.text = recognizedText
        }
    }
}

override fun onError(error: Int) {val errorMessage = when (error) {
        SpeechRecognizer.ERROR_AUDIO -> "音频错误"
        SpeechRecognizer.ERROR_CLIENT -> "客户端错误"
        // 其他错误处理...
        else -> "未知错误: $error"
    }
    Toast.makeText(this, errorMessage, Toast.LENGTH_SHORT).show()}

private fun isOnline(): Boolean {val connectivityManager = getSystemService(Context.CONNECTIVITY_SERVICE) as ConnectivityManager
    val networkInfo = connectivityManager.activeNetworkInfo
    return networkInfo?.isConnected == true
}

fun startListening(view: View) {if (!isOnline()) {Toast.makeText(this, "请检查网络连接", Toast.LENGTH_SHORT).show()
        return
    }
    // 继续语音识别...
}

避免长时间持续监听
在后台服务中使用PARTIAL_WAKE_LOCK
实现 JobIntentService 处理长时间任务

override fun onRequestPermissionsResult(
    requestCode: Int,
    permissions: Array<String>,
    grantResults: IntArray
) {super.onRequestPermissionsResult(requestCode, permissions, grantResults)

    when (requestCode) {
        REQUEST_RECORD_AUDIO -> {if (grantResults.isNotEmpty() && grantResults[0] == PackageManager.PERMISSION_GRANTED) {startListening()
            } else {showPermissionExplanationDialog()
            }
        }
    }
}

val intent = Intent(RecognizerIntent.ACTION_RECOGNIZE_SPEECH).apply {putExtra(RecognizerIntent.EXTRA_LANGUAGE, "zh-CN") // 中文
    // putExtra(RecognizerIntent.EXTRA_LANGUAGE, "en-US") // 英文
}

/app
  ├── src/main
  │   ├── java/com/example/speechrecognition
  │   │   ├── SpeechActivity.kt
  │   ├── res
  │   │   ├── layout/activity_speech.xml
  │   │   ├── values/strings.xml
  ├── AndroidManifest.xml