Question How can you compare words (Speech Recognizer)?

my code with windows.speech, but I want to use it for android:

Code (CSharp):

1. using UnityEngine;
2. using UnityEngine.Windows.Speech;
3.  
4. public class Spieler_WoerterErkennung : MonoBehaviour
5. {
6.     private AudioSource Audioquelle;
7.  
8.     [SerializeField]
9.     private string[] m_Keywords;
10.     private KeywordRecognizer m_Recognizer;
11.     private int zufall;
12.  
13.     void Start()
14.     {
15.         Audioquelle = GetComponent<AudioSource>();
16.  
17.         m_Recognizer = new KeywordRecognizer(m_Keywords);
18.         m_Recognizer.OnPhraseRecognized += OnPhraseRecognized;
19.         m_Recognizer.Start();
20.     }
21.  
22.     private void OnPhraseRecognized(PhraseRecognizedEventArgs args)
23.     {
24.         if (!Audioquelle.isPlaying)
25.         {
26.             // If the player says the same words like in Array '0', this if will be executed:
27.             if (args.text == m_Keywords[0])
28.             {
29.         Debug.Log("the same words");
30.             }
31.         }
32.  
33.     }
34.  
35.  
36.  
37.  
38. }

So far I have only worked with the Windows Speech Rechonizer in Unity, but did not consider that it might not work as soon as I try to export the application as an Android apk. that's why I've been looking for days and only found this video as a half-hearted solution. everything else is otherwise too expensive or too complicated. Is there a way how I can match words as templates in an if, as I have already done with Windows.speech? Because when I say the words: "Hellow how are you" into the microphone, an if should execute a Debug.Log to show that it has understood the words as these. But how do I do that if I can't use windows.speech because I'm making an app for android? I tried to orientate myself on this video, but it does not show in the video how to query the spoken words as an if with a prepared sentence by me, with which the spoken words are compared so that if they match the Debug.Log command is executed.

 

I suspect all this sorta "soft learning ML AI type stuff" is just going to the cloud.

There's probably a way with making a Google app on their back end, something to use their speech-to-text features.

I imagine you'd capture the audio, blast it to Google, they'd send you back some text.

 

Can you recommend me a Unity Asset Store Asset with this feature to make a speek regognisation and compare it in a if Statement of the same words?

 

I would start by looking in Google. Conveniently you can do so yourself, leaving me free to continue. Good luck!