Update app.py

app.py

@@ -6,6 +6,7 @@ import subprocess
 import gradio as gr
 import uuid
 from dotenv import load_dotenv
+from urllib.parse import urlparse
 
 # Load environment variables
 load_dotenv()
@@ -131,23 +132,63 @@ def combine_audio_video(video_path, audio_path, output_path):
 
     subprocess.run(cmd, check=True)
 
-def process_video(voice, video_url, text, progress=gr.Progress()):
-    session_id = str(uuid.uuid4())  # Generate a unique session ID
+def is_image_url(url):
+    parsed = urlparse(url)
+    path = parsed.path.lower()
+    return path.endswith(('.png', '.jpg', '.jpeg', '.gif', '.bmp'))
+
+def create_video_from_image(image_url, output_path, duration=10):
+    # Download the image
+    response = requests.get(image_url)
+    if response.status_code != 200:
+        raise Exception("Failed to download the image")
+    
+    temp_image_path = f"temp_image_{uuid.uuid4()}.jpg"
+    with open(temp_image_path, 'wb') as f:
+        f.write(response.content)
+    
+    # Create a 10-second video from the image
+    cmd = [
+        'ffmpeg', '-loop', '1', '-i', temp_image_path,
+        '-c:v', 'libx264', '-t', str(duration), '-pix_fmt', 'yuv420p',
+        '-vf', 'scale=1920:1080:force_original_aspect_ratio=decrease,pad=1920:1080:(ow-iw)/2:(oh-ih)/2',
+        '-y', output_path
+    ]
+    subprocess.run(cmd, check=True)
+    
+    # Clean up the temporary image file
+    os.remove(temp_image_path)
+    
+    return output_path
+
+def process_video(voice, media_url, text, progress=gr.Progress()):
+    session_id = str(uuid.uuid4())
     progress(0, desc="Generating speech...")
     audio_path = text_to_speech(voice, text, session_id)
     if not audio_path:
         return None, "Failed to generate speech audio."
     
-    progress(0.2, desc="Processing video...")
+    progress(0.2, desc="Processing media...")
     
     try:
-        progress(0.3, desc="Uploading audio...")
+        if is_image_url(media_url):
+            progress(0.3, desc="Converting image to video...")
+            video_path = f"temp_video_{session_id}.mp4"
+            create_video_from_image(media_url, video_path)
+            progress(0.4, desc="Uploading converted video...")
+            video_url = upload_file(video_path)
+            if not video_url:
+                raise Exception("Failed to upload converted video")
+        else:
+            video_url = media_url
+        
+        progress(0.5, desc="Uploading audio...")
         audio_url = upload_file(audio_path)
         
         if not audio_url:
             raise Exception("Failed to upload audio file")
         
-        progress(0.4, desc="Initiating lipsync...")
+        progress(0.6, desc="Initiating lipsync...")
         job_data = lipsync_api_call(video_url, audio_url)
         
         if "error" in job_data or "message" in job_data:
@@ -155,7 +196,7 @@ def process_video(voice, video_url, text, progress=gr.Progress()):
         
         job_id = job_data["id"]
         
-        progress(0.5, desc="Processing lipsync...")
+        progress(0.7, desc="Processing lipsync...")
         result_url = check_job_status(job_id)
         
         if result_url:
@@ -172,11 +213,12 @@ def process_video(voice, video_url, text, progress=gr.Progress()):
     except Exception as e:
         progress(0.8, desc="Falling back to simple combination...")
         try:
-            # Download the video from the URL
-            video_response = requests.get(video_url)
-            video_path = f"temp_video_{session_id}.mp4"
-            with open(video_path, "wb") as f:
-                f.write(video_response.content)
+            if 'video_path' not in locals():
+                # Download the video from the URL if it wasn't created from an image
+                video_response = requests.get(video_url)
+                video_path = f"temp_video_{session_id}.mp4"
+                with open(video_path, "wb") as f:
+                    f.write(video_response.content)
             
             output_path = f"output_{session_id}.mp4"
             combine_audio_video(video_path, audio_path, output_path)
@@ -199,22 +241,22 @@ def create_interface():
         with gr.Row():
             with gr.Column():
                 voice_dropdown = gr.Dropdown(choices=[v[0] for v in voices], label="Select Voice", value=voices[0][0] if voices else None)
-                video_url_input = gr.Textbox(label="Enter Video URL")
+                media_url_input = gr.Textbox(label="Enter Video or Image URL")
                 text_input = gr.Textbox(label="Enter text", lines=3)
                 generate_btn = gr.Button("Generate Video")
             with gr.Column():
                 video_output = gr.Video(label="Generated Video")
                 status_output = gr.Textbox(label="Status", interactive=False)
         
-        def on_generate(voice_name, video_url, text):
+        def on_generate(voice_name, media_url, text):
             voice_id = next((v[1] for v in voices if v[0] == voice_name), None)
             if not voice_id:
                 return None, "Invalid voice selected."
-            return process_video(voice_id, video_url, text)
+            return process_video(voice_id, media_url, text)
         
         generate_btn.click(
             fn=on_generate,
-            inputs=[voice_dropdown, video_url_input, text_input],
+            inputs=[voice_dropdown, media_url_input, text_input],
             outputs=[video_output, status_output]
         )