diff --git a/imaginairy/api/upscale.py b/imaginairy/api/upscale.py
index 83694672..16c39fb6 100644
--- a/imaginairy/api/upscale.py
+++ b/imaginairy/api/upscale.py
@@ -1,18 +1,21 @@
-from PIL import Image
+from typing import TYPE_CHECKING, Union
 
 from imaginairy.config import DEFAULT_UPSCALE_MODEL
-from imaginairy.enhancers.upscale import upscale_image
-from imaginairy.schema import LazyLoadingImage
+
+if TYPE_CHECKING:
+    from PIL import Image
+
+    from imaginairy.schema import LazyLoadingImage
 
 
 def upscale(
-    img: LazyLoadingImage | Image.Image | str,
+    img: "Union[LazyLoadingImage, Image.Image, str]",
     upscale_model: str = DEFAULT_UPSCALE_MODEL,
     tile_size: int = 512,
     tile_pad: int = 50,
     repetition: int = 1,
     device=None,
-) -> Image.Image:
+) -> "Image.Image":
     """
     Upscales an image using a specified super-resolution model.
 
@@ -31,6 +34,11 @@ def upscale(
     Returns:
         Image.Image: The upscaled image as a PIL Image object.
     """
+    from PIL import Image
+
+    from imaginairy.enhancers.upscale import upscale_image
+    from imaginairy.schema import LazyLoadingImage
+
     if isinstance(img, str):
         if img.startswith("https://"):
             img = LazyLoadingImage(url=img)
diff --git a/imaginairy/enhancers/upscale.py b/imaginairy/enhancers/upscale.py
index 2734526e..e076a34f 100644
--- a/imaginairy/enhancers/upscale.py
+++ b/imaginairy/enhancers/upscale.py
@@ -1,11 +1,15 @@
 import logging
-
-from PIL import Image
+from typing import TYPE_CHECKING, Union
 
 from imaginairy.config import DEFAULT_UPSCALE_MODEL
-from imaginairy.schema import LazyLoadingImage
 from imaginairy.utils import get_device
 
+if TYPE_CHECKING:
+    from PIL import Image
+
+    from imaginairy.schema import LazyLoadingImage
+
+
 upscale_model_lookup = {
     # RealESRGAN
     "ultrasharp": "https://huggingface.co/lokCX/4x-Ultrasharp/resolve/1856559b50de25116a7c07261177dd128f1f5664/4x-UltraSharp.pth",
@@ -22,13 +26,13 @@
 
 
 def upscale_image(
-    img: LazyLoadingImage | Image.Image,
+    img: "Union[LazyLoadingImage, Image.Image]",
     upscaler_model: str = DEFAULT_UPSCALE_MODEL,
     tile_size: int = 512,
     tile_pad: int = 50,
     repetition: int = 1,
     device=None,
-) -> Image.Image:
+) -> "Image.Image":
     """
     Upscales an image using a specified super-resolution model.
 
@@ -95,12 +99,14 @@ def upscale_image(
     return image
 
 
-def load_image(img: LazyLoadingImage | Image.Image):
+def load_image(img: "Union[LazyLoadingImage, Image.Image]"):
     """
     Converts a LazyLoadingImage or PIL Image into a PyTorch tensor.
     """
     from torchvision import transforms
 
+    from imaginairy.schema import LazyLoadingImage
+
     if isinstance(img, LazyLoadingImage):
         img = img.as_pillow()
     transform = transforms.ToTensor()
diff --git a/imaginairy/utils/__init__.py b/imaginairy/utils/__init__.py
index fc9284a9..d13b76bc 100644
--- a/imaginairy/utils/__init__.py
+++ b/imaginairy/utils/__init__.py
@@ -240,10 +240,9 @@ def glob_expand_paths(paths):
             p = os.path.expanduser(p)
             if os.path.exists(p) and os.path.isfile(p):
                 expanded_paths.append(p)
-            elif os.path.exists(p) and os.path.isdir(p):
-                expanded_paths.extend(glob.glob(os.path.expanduser(p)))
             else:
-                print(f"Warning: {p} does not exist.")
+                expanded_paths.extend(glob.glob(os.path.expanduser(p)))
+
     return expanded_paths
 
 
diff --git a/imaginairy/utils/tile_up.py b/imaginairy/utils/tile_up.py
index 50ba7b42..cc1ec779 100644
--- a/imaginairy/utils/tile_up.py
+++ b/imaginairy/utils/tile_up.py
@@ -1,19 +1,21 @@
 import logging
 import math
+from typing import TYPE_CHECKING
 
-import torch
-from torch import Tensor
+if TYPE_CHECKING:
+    import torch
+    from torch import Tensor
 
 logger = logging.getLogger(__name__)
 
 
 def tile_process(
-    img: Tensor,
+    img: "Tensor",
     scale: int,
-    model: torch.nn.Module,
+    model: "torch.nn.Module",
     tile_size: int = 512,
     tile_pad: int = 50,
-) -> Tensor:
+) -> "Tensor":
     """
     Process an image by tiling it, processing each tile, and then merging them back into one image.
 
@@ -27,6 +29,8 @@ def tile_process(
     Returns:
     Tensor: The processed output image.
     """
+    import torch
+
     batch, channel, height, width = img.shape
     output_height = height * scale
     output_width = width * scale