Is there a way to use the SpeechRecognizer API directly for speech input?

Можно сделать это как:

import android.app.Activity
import androidx.appcompat.app.AppCompatActivity
import android.os.Bundle
import kotlinx.android.synthetic.main.activity_main.*
import android.widget.Toast
import android.content.ActivityNotFoundException
import android.speech.RecognizerIntent
import android.content.Intent

class MainActivity : AppCompatActivity() {
    private val REQ_CODE = 100

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

        speak.setOnClickListener {
            val intent = Intent(RecognizerIntent.ACTION_RECOGNIZE_SPEECH)
            intent.putExtra(RecognizerIntent.EXTRA_LANGUAGE_MODEL,
                    RecognizerIntent.LANGUAGE_MODEL_FREE_FORM)
            intent.putExtra(RecognizerIntent.EXTRA_LANGUAGE,  "ar-JO") //  Locale.getDefault()
            intent.putExtra(RecognizerIntent.EXTRA_PROMPT, "Need to speak")
            try {
                startActivityForResult(intent, REQ_CODE)
            } catch (a: ActivityNotFoundException) {
                Toast.makeText(applicationContext,
                        "Sorry your device not supported",
                        Toast.LENGTH_SHORT).show()
            }
        }
    }

    override fun onActivityResult(requestCode: Int, resultCode: Int, data: Intent?) {
        super.onActivityResult(requestCode, resultCode, data)

        when (requestCode) {
            REQ_CODE -> {
                if (resultCode == Activity.RESULT_OK && data != null) {
                    val result = data
                            .getStringArrayListExtra(RecognizerIntent.EXTRA_RESULTS)
                    println("result: $result")
                    text.text = result[0]
                }
            }
        }
    }
}

Эти layout мог быть просто:

<?xml version = "1.0" encoding = "utf-8"?>
<RelativeLayout xmlns:android = "http://schemas.android.com/apk/res/android"
    xmlns:app = "http://schemas.android.com/apk/res-auto"
    xmlns:tools = "http://schemas.android.com/tools"
    android:layout_width = "match_parent"
    android:layout_height = "match_parent"
    tools:context = ".MainActivity">
    <LinearLayout
        android:layout_width = "match_parent"
        android:gravity = "center"
        android:layout_height = "match_parent">
        <TextView
            android:id = "@+id/text"
            android:textSize = "30sp"
            android:layout_width = "wrap_content"
            android:layout_height = "wrap_content"/>
    </LinearLayout>
    <LinearLayout
        android:layout_width = "wrap_content"
        android:layout_alignParentBottom = "true"
        android:layout_centerInParent = "true"
        android:orientation = "vertical"
        android:layout_height = "wrap_content">
        <ImageView
            android:id = "@+id/speak"
            android:layout_width = "wrap_content"
            android:layout_height = "wrap_content"
            android:background = "?selectableItemBackground"
            android:src = "@android:drawable/ic_btn_speak_now"/>
    </LinearLayout>
</RelativeLayout>

другой способ, о котором Вы спрашиваете, немного длиннее, но дайте Вам больше контроля, также не делает раздражающий Вы с диалогом Google Assistance:

1-Первых Вам нужно к главным полномочиям в Manifest файл:

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

2-я консолидирую все ответы выше, как:

• Создают RecognitionListener класс, как:

private val TAG = "Driver-Assistant"

class Listener(context: Context): RecognitionListener {
    private var ctx = context

    override fun onReadyForSpeech(params: Bundle?) {
        Log.d(TAG, "onReadyForSpeech")
    }

    override fun onRmsChanged(rmsdB: Float) {
        Log.d(TAG, "onRmsChanged")
    }

    override fun onBufferReceived(buffer: ByteArray?) {
        Log.d(TAG, "onBufferReceived")
    }

    override fun onPartialResults(partialResults: Bundle?) {
        Log.d(TAG, "onPartialResults")
    }

    override fun onEvent(eventType: Int, params: Bundle?) {
        Log.d(TAG, "onEvent")
    }

    override fun onBeginningOfSpeech() {
        Toast.makeText(ctx, "Speech started", Toast.LENGTH_LONG).show()
    }

    override fun onEndOfSpeech() {
        Toast.makeText(ctx, "Speech finished", Toast.LENGTH_LONG).show()
    }

    override fun onError(error: Int) {
        var string = when (error) {
            6 -> "No speech input"
            4 -> "Server sends error status"
            8 -> "RecognitionService busy."
            7 -> "No recognition result matched."
            1 -> "Network operation timed out."
            2 -> "Other network related errors."
            9 -> "Insufficient permissions"
            5 -> " Other client side errors."
            3 -> "Audio recording error."
            else -> "unknown!!"
        }
        Toast.makeText(ctx, "sorry error occurred: $string", Toast.LENGTH_LONG).show()
    }

    override fun onResults(results: Bundle?) {
        Log.d(TAG, "onResults $results")
        val data = results!!.getStringArrayList(SpeechRecognizer.RESULTS_RECOGNITION)
        display.text = data!![0]
    }
}

• В основном файле Вы должны определить SpeechRecognizer, добавив вышеупомянутое listner к нему, и не забываете просить разрешение во время выполнения, все вместе ниже:

lateinit var sr: SpeechRecognizer
lateinit var display: TextView

class MainActivity : AppCompatActivity() {

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

        display = text

        if (ContextCompat.checkSelfPermission(this,
                        Manifest.permission.RECORD_AUDIO)
                != PackageManager.PERMISSION_GRANTED) {
            if (ActivityCompat.shouldShowRequestPermissionRationale(this,
                            Manifest.permission.RECORD_AUDIO)) {
            } else {
                ActivityCompat.requestPermissions(this,
                        arrayOf(Manifest.permission.RECORD_AUDIO),
                        527)
            }
        }

        sr = SpeechRecognizer.createSpeechRecognizer(this)
        sr.setRecognitionListener(Listener(this))

        speak.setOnClickListener {
            val intent = Intent(RecognizerIntent.ACTION_RECOGNIZE_SPEECH)
            intent.putExtra(RecognizerIntent.EXTRA_LANGUAGE_MODEL,
                    RecognizerIntent.LANGUAGE_MODEL_FREE_FORM)
            intent.putExtra(RecognizerIntent.EXTRA_LANGUAGE,  "ar-JO") //  Locale.getDefault()
            sr.startListening(intent)
        }

    }

    override fun onRequestPermissionsResult(requestCode: Int, permissions: Array<out String>, grantResults: IntArray) {
        super.onRequestPermissionsResult(requestCode, permissions, grantResults)
        when (requestCode) {
            527  -> if (grantResults.isNotEmpty()
                    && grantResults[0] == PackageManager.PERMISSION_GRANTED) {

                Toast.makeText(this, "Permission granted", Toast.LENGTH_SHORT).show()
            } else {
                Toast.makeText(this, "Permission not granted", Toast.LENGTH_SHORT).show()
            }
        }
    }
}