Update app.py

app.py

@@ -3,13 +3,17 @@ import edge_tts
 import asyncio
 import tempfile
 
+from expressive_tts import expressive_tts
+
 
 # -----------------------------
-# Voice Loader
+# Voice Loader (Baseline)
 # -----------------------------
 
 async def get_voices():
+
     voices = await edge_tts.list_voices()
+
     return {
         f"{v['ShortName']} - {v['Locale']} ({v['Gender']})": v["ShortName"]
         for v in voices
@@ -17,7 +21,7 @@ async def get_voices():
 
 
 # -----------------------------
-# TTS Engine
+# Baseline TTS (Gen-0)
 # -----------------------------
 
 async def text_to_speech(text, voice, rate, pitch):
@@ -44,22 +48,45 @@ async def text_to_speech(text, voice, rate, pitch):
         delete=False,
         suffix=".mp3",
     ) as tmp:
+
         path = tmp.name
+
         await communicate.save(path)
 
     return path, None
 
 
-async def tts_interface(text, voice, rate, pitch):
+# -----------------------------
+# Combined Interface
+# -----------------------------
+
+async def tts_interface(
+    text,
+    voice,
+    rate,
+    pitch,
+    emotion
+):
 
-    audio, warning = await text_to_speech(
+    # Gen-0 TTS
+    normal_audio, warning = await text_to_speech(
         text, voice, rate, pitch
     )
 
     if warning:
-        return audio, gr.Warning(warning)
+        return None, None, gr.Warning(warning)
 
-    return audio, None
+    # Gen-2 TTS (Expressive)
+    expressive_audio, err = expressive_tts(
+        text=text,
+        emotion=emotion,
+        output_path="expressive.wav"
+    )
+
+    if err:
+        return normal_audio, None, gr.Warning(err)
+
+    return normal_audio, expressive_audio, None
 
 
 # -----------------------------
@@ -73,52 +100,92 @@ async def create_demo():
     with gr.Blocks(analytics_enabled=False) as demo:
 
         gr.Markdown("""
-        # 🎙️ AI Text-to-Speech
+        # 🎙️ AI Text-to-Speech Lab
+
+        **Gen-0 vs Gen-2 Comparison**
 
-        Convert text into natural speech using Microsoft Edge voices.
+        - Left: Traditional flat TTS (Edge)
+        - Right: Emotional AI TTS (XTTS-v2)
         """)
 
+
+        # -----------------------------
+        # Inputs
+        # -----------------------------
+
         text_input = gr.Textbox(
             label="Text",
             lines=6,
-            placeholder="Paste lecture notes, narration, or scripts here...",
+            placeholder="Paste lecture notes, narration, or scripts here..."
         )
 
+
         voice_dropdown = gr.Dropdown(
             choices=[""] + list(voices.keys()),
-            label="Voice",
-            value="",
+            label="Baseline Voice",
+            value=""
         )
 
+
+        emotion_dropdown = gr.Dropdown(
+            choices=[
+                "neutral",
+                "happy",
+                "sad",
+                "angry",
+                "calm",
+                "excited"
+            ],
+            value="neutral",
+            label="AI Emotion Style"
+        )
+
+
         with gr.Row():
 
             rate_slider = gr.Slider(
                 -50, 50, 0,
                 step=1,
-                label="Rate (%)",
+                label="Rate (%)"
             )
 
             pitch_slider = gr.Slider(
                 -20, 20, 0,
                 step=1,
-                label="Pitch (Hz)",
+                label="Pitch (Hz)"
             )
 
 
         generate_btn = gr.Button(
             "Generate Audio",
-            variant="primary",
+            variant="primary"
         )
 
 
-        audio_output = gr.Audio(
-            label="Output",
-            type="filepath",
-        )
+        # -----------------------------
+        # Outputs
+        # -----------------------------
+
+        with gr.Row():
+
+            normal_audio = gr.Audio(
+                label="Gen-0: Flat TTS (Edge)",
+                type="filepath"
+            )
+
+            expressive_audio = gr.Audio(
+                label="Gen-2: Emotional AI TTS",
+                type="filepath"
+            )
+
 
         warning_md = gr.Markdown(visible=False)
 
 
+        # -----------------------------
+        # Events
+        # -----------------------------
+
         generate_btn.click(
             fn=tts_interface,
             inputs=[
@@ -126,11 +193,13 @@ async def create_demo():
                 voice_dropdown,
                 rate_slider,
                 pitch_slider,
+                emotion_dropdown
             ],
             outputs=[
-                audio_output,
-                warning_md,
-            ],
+                normal_audio,
+                expressive_audio,
+                warning_md
+            ]
         )
 
 
@@ -145,13 +214,12 @@ async def main():
 
     demo = await create_demo()
 
-    demo.queue(default_concurrency_limit=50)
+    demo.queue(default_concurrency_limit=20)
 
     demo.launch(
         theme=gr.themes.Base()
     )
 
 
-
 if __name__ == "__main__":
     asyncio.run(main())