Update app.py

app.py

@@ -27,26 +27,26 @@ def get_video_res(img_path, audio_path, res_video_path, dynamic_scale=1.0):
     expand_ratio = 0.5
     min_resolution = 512
     inference_steps = 25
-    
+
     # Get audio duration
     audio = AudioSegment.from_file(audio_path)
     duration = len(audio) / 1000.0  # Convert to seconds
-    
+
     face_info = pipe.preprocess(img_path, expand_ratio=expand_ratio)
     print(f"Face detection info: {face_info}")
     print(f"Audio duration: {duration} seconds")
-    
+
     if face_info['face_num'] > 0:
         crop_image_path = img_path + '.crop.png'
         pipe.crop_image(img_path, crop_image_path, face_info['crop_bbox'])
         img_path = crop_image_path
         os.makedirs(os.path.dirname(res_video_path), exist_ok=True)
-        
+
         # Process with full audio duration
         pipe.process(
-            img_path, 
-            audio_path, 
-            res_video_path, 
+            img_path,
+            audio_path,
+            res_video_path,
             min_resolution=min_resolution,
             inference_steps=inference_steps,
             dynamic_scale=dynamic_scale,
@@ -57,8 +57,8 @@ def get_video_res(img_path, audio_path, res_video_path, dynamic_scale=1.0):
 
 tmp_path = './tmp_path/'
 res_path = './res_path/'
-os.makedirs(tmp_path, exist_ok=1)
-os.makedirs(res_path, exist_ok=1)
+os.makedirs(tmp_path, exist_ok=True)
+os.makedirs(res_path, exist_ok=True)
 
 def process_sonic(image, audio, dynamic_scale):
     # Input validation
@@ -66,15 +66,15 @@ def process_sonic(image, audio, dynamic_scale):
         raise gr.Error("Please upload an image")
     if audio is None:
         raise gr.Error("Please upload an audio file")
-        
+
     img_md5 = get_md5(np.array(image))
     audio_md5 = get_md5(audio[1])
     print(f"Processing with image hash: {img_md5}, audio hash: {audio_md5}")
-    
+
     sampling_rate, arr = audio[:2]
     if len(arr.shape) == 1:
         arr = arr[:, None]
-    
+
     # Create audio segment
     audio_segment = AudioSegment(
         arr.tobytes(),
@@ -83,18 +83,18 @@ def process_sonic(image, audio, dynamic_scale):
         channels=arr.shape[1]
     )
     audio_segment = audio_segment.set_frame_rate(sampling_rate)
-    
+
     # Generate paths
     image_path = os.path.abspath(os.path.join(tmp_path, f'{img_md5}.png'))
     audio_path = os.path.abspath(os.path.join(tmp_path, f'{audio_md5}.wav'))
     res_video_path = os.path.abspath(os.path.join(res_path, f'{img_md5}_{audio_md5}_{dynamic_scale}.mp4'))
-    
+
     # Save inputs if they don't exist
     if not os.path.exists(image_path):
         image.save(image_path)
     if not os.path.exists(audio_path):
         audio_segment.export(audio_path, format="wav")
-    
+
     # Process or return cached result
     if os.path.exists(res_video_path):
         print(f"Using cached result: {res_video_path}")
@@ -103,6 +103,11 @@ def process_sonic(image, audio, dynamic_scale):
         print(f"Generating new video with dynamic scale: {dynamic_scale}")
         return get_video_res(image_path, audio_path, res_video_path, dynamic_scale)
 
+# Dummy get_example function to prevent errors if examples are not defined
+def get_example():
+    # 예시가 없다면 빈 리스트를 반환하거나 실제 예시 데이터를 입력할 수 있습니다.
+    return []
+
 # Enhanced UI
 css = """
 .gradio-container {
@@ -131,22 +136,21 @@ with gr.Blocks(css=css) as demo:
             <p>Transform still images into dynamic videos synchronized with audio</p>
         </div>
     """)
-    
+
     with gr.Row():
         with gr.Column():
             image_input = gr.Image(
                 type='pil',
                 label="Portrait Image",
-                elem_id="image_input",
-                tool="select"
+                elem_id="image_input"
             )
-            
+
             audio_input = gr.Audio(
                 label="Voice/Audio Input",
                 elem_id="audio_input",
                 type="numpy"
             )
-            
+
             with gr.Box(elem_classes="parameter-section"):
                 dynamic_scale = gr.Slider(
                     minimum=0.5,
@@ -156,19 +160,19 @@ with gr.Blocks(css=css) as demo:
                     label="Animation Intensity",
                     info="Adjust to control movement intensity (0.5: subtle, 2.0: dramatic)"
                 )
-                
+
             process_btn = gr.Button(
-                "Generate Animation", 
+                "Generate Animation",
                 variant="primary",
                 elem_id="process_btn"
             )
-            
+
         with gr.Column():
             video_output = gr.Video(
                 label="Generated Animation",
                 elem_id="video_output"
             )
-            
+
     # Process button click
     process_btn.click(
         fn=process_sonic,
@@ -176,7 +180,7 @@ with gr.Blocks(css=css) as demo:
         outputs=video_output,
         api_name="animate"
     )
-    
+
     # Examples section
     gr.Examples(
         examples=get_example(),
@@ -186,7 +190,7 @@ with gr.Blocks(css=css) as demo:
         cache_examples=False,
         elem_classes="example-section"
     )
-    
+
     # Footer with attribution and links
     gr.HTML("""
         <div style="text-align: center; margin-top: 2em;">
@@ -202,4 +206,4 @@ with gr.Blocks(css=css) as demo:
         </div>
     """)
 
-demo.launch()
+demo.launch()