diff --git a/dan/utils.py b/dan/utils.py
index 6ceede28d5b1870f7e527e43d07b3536a3227e61..995465a734bdb2c467a1075794e3c5d5f49d56ec 100644
--- a/dan/utils.py
+++ b/dan/utils.py
@@ -24,27 +24,23 @@ def pad_sequences_1D(data, padding_value):
     return padded_data
 
 
-def pad_images(data):
+def pad_images(images):
     """
-    Pad the images so that they are in the middle of the large padded image (tb-lr mode).
-    :param data: List of numpy arrays.
-    :return padded_data: A tensor containing all the padded images.
+    Pad the images so that they are at the top left of the large padded image.
+    :param images: List of images as torch tensors.
+    :return padded_images: A tensor containing all the padded images.
     """
-    longest_x = max([x.shape[0] for x in data])
-    longest_y = max([x.shape[1] for x in data])
-    padded_data = np.zeros((len(data), longest_x, longest_y, data[0].shape[2]))
-    for index, image in enumerate(data):
-        delta_x = longest_x - image.shape[0]
-        delta_y = longest_y - image.shape[1]
-        top, bottom = delta_x // 2, delta_x - (delta_x // 2)
-        left, right = delta_y // 2, delta_y - (delta_y // 2)
-        padded_data[
+    longest_x = max([x.shape[0] for x in images])
+    longest_y = max([x.shape[1] for x in images])
+    padded_images = np.zeros((len(images), longest_x, longest_y, images[0].shape[2]))
+    for index, image in enumerate(images):
+        padded_images[
             index,
-            top : padded_data.shape[1] - bottom,
-            left : padded_data.shape[2] - right,
+            0 : image.shape[0],
+            0 : image.shape[1],
             :,
         ] = image
-    return padded_data
+    return padded_images
 
 
 def read_image(filename, scale=1.0):