Started to work on voice selection for speech synthesis

Assets/Scripts/Services/Speech/Microsoft/WindowsSpeechSynthesis2.cs

@@ -0,0 +1,176 @@
+#if UNITY_STANDALONE_WIN || UNITY_EDITOR_WIN
+#define MS_SPEECH_SYNTHESIS
+#endif
+
+using System;
+using System.Collections;
+using System.Collections.Generic;
+using System.IO;
+using UnityEngine;
+
+#if MS_SPEECH_SYNTHESIS
+using System.Runtime.InteropServices;
+#endif
+
+namespace Demonixis.InMoov.Services.Speech
+{
+    [RequireComponent(typeof(AudioSource))]
+    public class WindowsSpeechSynthesis2 : SpeechSynthesisService
+    {
+        const int TRUE = 1, FALSE = 0;
+
+        [DllImport("TTSDLL", EntryPoint = "CreateTTS")]
+        static extern IntPtr CreateTTS();
+
+        [DllImport("TTSDLL", EntryPoint = "GetVoiceCount")]
+        static extern int GetVoiceCount(IntPtr tts);
+
+        [DllImport("TTSDLL", EntryPoint = "GetNowVoice", CallingConvention = CallingConvention.Cdecl)]
+        static extern IntPtr GetNowVoice(IntPtr tts);
+
+        [DllImport("TTSDLL", EntryPoint = "GetVoiceByNum", CallingConvention = CallingConvention.Cdecl)]
+        static extern IntPtr GetVoiceByNum(IntPtr tts, int num);
+
+        [DllImport("TTSDLL", EntryPoint = "GetErrorMessage", CallingConvention = CallingConvention.Cdecl)]
+        static extern IntPtr GetErrorMessage(IntPtr tts);
+
+        [DllImport("TTSDLL", EntryPoint = "SetVoiceByName", CallingConvention = CallingConvention.Cdecl)]
+        static extern int SetVoiceByName(IntPtr tts, [MarshalAs(UnmanagedType.LPStr)] string name);
+
+        [DllImport("TTSDLL", EntryPoint = "SetVoiceByNum", CallingConvention = CallingConvention.Cdecl)]
+        static extern int SetVoiceByNum(IntPtr tts, int num);
+
+        [DllImport("TTSDLL", EntryPoint = "SetOutputPath", CallingConvention = CallingConvention.Cdecl)]
+        static extern int SetOutputPath(IntPtr tts, [MarshalAs(UnmanagedType.LPStr)] string filePath);
+
+
+        [DllImport("TTSDLL", EntryPoint = "TextToSpeech", CallingConvention = CallingConvention.Cdecl)]
+        static extern int TextToSpeech(IntPtr tts, [MarshalAs(UnmanagedType.LPStr)] string text, bool output = false);
+
+        private IntPtr _speechHandle;
+        private AudioSource _audioSource;
+        private string _ttsFilePath = Application.streamingAssetsPath + "/";
+        private string _ttsFileName = "tts_cache.wav";
+
+        private void Awake()
+        {
+            _audioSource = gameObject.AddComponent<AudioSource>();
+            _speechHandle = CreateTTS();
+            SetPath(_ttsFilePath, _ttsFileName);
+        }
+
+        public string GetVoice(int num = -1)
+        {
+            IntPtr ptr;
+            string str;
+            if (num < 0)
+                ptr = GetNowVoice(_speechHandle);
+            else
+                ptr = GetVoiceByNum(_speechHandle, num);
+            str = Marshal.PtrToStringUni(ptr);
+            Marshal.FreeHGlobal(ptr);
+            return str;
+        }
+
+        public string[] GetVoiceNames()
+        {
+            var count = GetVoiceCount(_speechHandle);
+            var names = new List<string>();
+            IntPtr ptr;
+
+            for (var i = 0; i < count; i++)
+            {
+                ptr = GetVoiceByNum(_speechHandle, i);
+                names.Add(Marshal.PtrToStringUni(ptr));
+                Marshal.FreeHGlobal(ptr);
+            }
+
+            return names.ToArray();
+        }
+
+        public void Speech(string text)
+        {
+            if (TextToSpeech(_speechHandle, text, true) == 0)
+            {
+                Debug.Log(GetMessage());
+            }
+
+            PlayTTS();
+        }
+
+        private void SetPath(string path, string file)
+        {
+            if (!Directory.Exists(path))
+            {
+                Directory.CreateDirectory(path);
+            }
+
+            if (SetOutputPath(_speechHandle, path + file) == FALSE)
+            {
+                Debug.Log("fail to set output:" + GetMessage());
+            }
+        }
+
+        public void SetVoice(int num)
+        {
+            if (num < 0) return;
+
+            if (SetVoiceByNum(_speechHandle, num) == FALSE)
+            {
+                Debug.Log(GetMessage());
+            }
+        }
+
+        public void SetVoice(string voiceName)
+        {
+            if (voiceName.Length <= 0) return;
+
+            if (SetVoiceByName(_speechHandle, voiceName) == FALSE)
+            {
+                Debug.Log(GetMessage());
+            }
+        }
+
+        private string GetMessage()
+        {
+            var handle = GetErrorMessage(_speechHandle);
+            var error = Marshal.PtrToStringUni(handle);
+            Marshal.FreeHGlobal(handle);
+            return error;
+        }
+
+        public void PlayTTS()
+        {
+            EnsureClipDisposed();
+
+            // Load Clip then assign to audio source and play
+            StartCoroutine(LoadClipCoroutine("file:///" + _ttsFilePath + _ttsFileName));
+        }
+
+        IEnumerator LoadClipCoroutine(string file)
+        {
+            var request = new WWW(file);
+            yield return request;
+
+            var clip = request.GetAudioClip(true, true);
+            _audioSource.clip = clip;
+            _audioSource.Play();
+
+            request.Dispose();
+        }
+
+        private void EnsureClipDisposed()
+        {
+            if (_audioSource.clip == null) return;
+
+            if (_audioSource.isPlaying)
+                _audioSource.Stop();
+
+            var clip = _audioSource.clip;
+            _audioSource.clip = null;
+
+            clip.UnloadAudioData();
+            Destroy(clip); // This is important to avoid memory leak
+        }
+    }
+}

Assets/Scripts/Services/Speech/VoiceRSS.cs

@@ -8,6 +8,18 @@ namespace Demonixis.InMoov.Services.Speech
     [RequireComponent(typeof(AudioSource))]
     public sealed class VoiceRSS : SpeechSynthesisService
     {
+        public enum Voices
+        {
+            enUS_Linda,
+            enUS_Amy,
+            enUS_Mary,
+            enUS_John,
+            frFR_Bette,
+            frFR_Iva,
+            frFR_Zola,
+            frFR_Axel
+        }
+
         private TextToSpeechManager _ttsManager;
         private AudioSource _audioSource;
 
@@ -53,6 +65,19 @@ public override void SetLanguage(string culture)
             // Feel free to continue this list ;)      
         }
 
+        public override void SelectVoice(string voiceName)
+        {
+            // voiceName is like frFR_VoiceName
+            var tmp = voiceName.Split('_');
+            if (tmp.Length != 2)
+            {
+                Debug.LogError($"[VoiceRSS] SelectVoice/Bad voice format: {voiceName}");
+                return;
+            }
+
+            _voice = tmp[1];
+        }
+
         public override void Speak(string text)
         {
             if (Paused) return;

Assets/Scripts/UI/Service/ServicePanel.cs

@@ -10,18 +10,24 @@
 
 public sealed class ServicePanel : MonoBehaviour
 {
+    [Header("Brain")]
     [SerializeField] private TMP_Dropdown _botServiceList;
     [SerializeField] private Toggle _botServiceStatus;
     [SerializeField] private TMP_Dropdown _speechRecognitionServiceList;
     [SerializeField] private Toggle _speechRecognitionServiceStatus;
     [SerializeField] private TMP_Dropdown _speechSynthesisServiceList;
+    [SerializeField] private TMP_Dropdown _speechSynthesisVoiceList;
     [SerializeField] private Toggle _speechSynthesisServiceStatus;
+
+    [Header("Body")]
     [SerializeField] private TMP_Dropdown _servoMixerServiceList;
     [SerializeField] private Toggle _servoMixerServiceStatus;
-    [SerializeField] private TMP_Dropdown _computerVisionServiceList;
-    [SerializeField] private Toggle _computerVisionServiceStatus;
     [SerializeField] private TMP_Dropdown _navigationServiceList;
     [SerializeField] private Toggle _navigationServiceStatus;
+
+    [Header("Perception")]
+    [SerializeField] private TMP_Dropdown _computerVisionServiceList;
+    [SerializeField] private Toggle _computerVisionServiceStatus;
 
     private void Start()
     {