Update app.py

app.py

@@ -69,13 +69,14 @@ def resize_image_to_bucket(image: Union[Image.Image, np.ndarray], bucket_reso: T
     """
     Resize the image to the bucket resolution.
     """
-    is_pil_image = isinstance(image, Image.Image)
-    if is_pil_image:
-        image_width, image_height = image.size
-    else:
-        image_height, image_width = image.shape[:2]
+    if isinstance(image, Image.Image):
+        image = np.array(image)
+    elif not isinstance(image, np.ndarray):
+        raise ValueError("Image must be a PIL Image or NumPy array")
+
+    image_height, image_width = image.shape[:2]
     if bucket_reso == (image_width, image_height):
-        return np.array(image) if is_pil_image else image
+        return image
     bucket_width, bucket_height = bucket_reso
     scale_width = bucket_width / image_width
     scale_height = bucket_height / image_height
@@ -83,32 +84,31 @@ def resize_image_to_bucket(image: Union[Image.Image, np.ndarray], bucket_reso: T
     image_width = int(image_width * scale + 0.5)
     image_height = int(image_height * scale + 0.5)
     if scale > 1:
-        image = Image.fromarray(image) if not is_pil_image else image
+        image = Image.fromarray(image)
         image = image.resize((image_width, image_height), Image.LANCZOS)
         image = np.array(image)
     else:
-        image = np.array(image) if is_pil_image else image
         image = cv2.resize(image, (image_width, image_height), interpolation=cv2.INTER_AREA)
     # crop the image to the bucket resolution
     crop_left = (image_width - bucket_width) // 2
     crop_top = (image_height - bucket_height) // 2
-    image = image[crop_top : crop_top + bucket_height, crop_left : crop_left + bucket_width]
+    image = image[crop_top:crop_top + bucket_height, crop_left:crop_left + bucket_width]
     return image
 
 
 def generate_video(pipe, prompt: str, frame1: Image.Image, frame2: Image.Image, guidance_scale: float, num_frames: int, num_inference_steps: int) -> bytes:
-    # Load and preprocess frames
-    # print(f"Frame 1 Path: {frame1_path}, Type: {type(frame1_path)}")
-    # print(f"Frame 2 Path: {frame2_path}, Type: {type(frame2_path)}")
+    # Debugging print statements
+    print(f"Frame 1 Type: {type(frame1)}")
+    print(f"Frame 2 Type: {type(frame2)}")
     
     # Load and preprocess frames
-    cond_frame1 = frame1
-    cond_frame2 = frame2
+    cond_frame1 = np.array(frame1)
+    cond_frame2 = np.array(frame2)
     height, width = 720, 1280
     cond_frame1 = resize_image_to_bucket(cond_frame1, bucket_reso=(width, height))
     cond_frame2 = resize_image_to_bucket(cond_frame2, bucket_reso=(width, height))
     cond_video = np.zeros(shape=(num_frames, height, width, 3))
-    cond_video[0], cond_video[-1] = np.array(cond_frame1), np.array(cond_frame2)
+    cond_video[0], cond_video[-1] = cond_frame1, cond_frame2
     cond_video = torch.from_numpy(cond_video.copy()).permute(0, 3, 1, 2)
     cond_video = torch.stack([video_transforms(x) for x in cond_video], dim=0).unsqueeze(0)
     with torch.no_grad():
@@ -324,7 +324,7 @@ def main():
     )
     
     # Launch the Gradio app
-    iface.launch()
+    iface.launch(show_error=True)
 
 if __name__ == "__main__":
     main()