Update app.py

app.py

@@ -1,6 +1,8 @@
 import os
+import zipfile
 import subprocess
 import torch
+import gradio as gr
 from tortoise.api import TextToSpeech
 from tortoise.utils.audio import save_wav
 
@@ -18,15 +20,10 @@ def setup_wav2lip():
     
     return None
 
-
 def generate_speech(text, voice_model, output_wav):
     """ Uses Tortoise TTS to generate speech. """
-    print(f"Generating speech for: {text} in {voice_model} voice...")
     tts = TextToSpeech()
-    voice = voice_model  # Example: 'trump'
-    
-    # Generate speech using Tortoise
-    speech = tts.tts(text, voice)
+    speech = tts.tts(text, voice_model)
     save_wav(speech, output_wav)
 
 def run_wav2lip(video_path, audio_path, output_video):
@@ -40,22 +37,44 @@ def run_wav2lip(video_path, audio_path, output_video):
     ]
     subprocess.run(command, check=True)
 
-def main():
-    setup_wav2lip()  # Ensure Wav2Lip is ready
+def process_lipsync(text, voice_model, video_file):
+    """ Handles text-to-speech generation and lip-syncing. """
+    error = setup_wav2lip()
+    if error:
+        return error, None
+    
+    if not text or not video_file:
+        return "Please provide both text and a video.", None
     
-    text = "Hello, this is a test speech."
-    voice_model = "trump"
     output_wav = "generated_speech.wav"
-    video_path = "trump_video.mp4"
     output_video = "lip_synced_output.mp4"
+    video_path = "input_video.mp4"
     
-    # Step 1: Generate Speech
+    # Save the uploaded video file
+    with open(video_path, "wb") as f:
+        f.write(video_file.read())
+
     generate_speech(text, voice_model, output_wav)
-    
-    # Step 2: Run Wav2Lip
     run_wav2lip(video_path, output_wav, output_video)
     
-    print(f"Lip-synced video saved as {output_video}")
+    return "Lip-synced video generated!", output_video
+
+def main():
+    iface = gr.Interface(
+        fn=process_lipsync,
+        inputs=[
+            gr.Textbox(label="Enter text for speech synthesis"),
+            gr.Dropdown(["trump", "elon", "obama"], label="Choose a voice model"),
+            gr.File(label="Upload a video (MP4)")
+        ],
+        outputs=[
+            gr.Textbox(label="Status"),
+            gr.Video(label="Generated Lip-Synced Video")
+        ],
+        title="TTS & Lip Sync Generator",
+        description="Enter text, select a voice, and upload a video to generate a lip-synced output."
+    )
+    iface.launch()
 
 if __name__ == "__main__":
     main()