diff --git a/test/test_datasets.py b/test/test_datasets.py
index 48d08b846de..2c74332a833 100644
--- a/test/test_datasets.py
+++ b/test/test_datasets.py
@@ -2746,15 +2746,15 @@ def inject_fake_data(self, tmpdir, config):
         datasets_utils.create_image_folder(
             root=split_dir,
             name="image_2",
-            file_name_fn=lambda i: f"{i:06d}_10.png",
-            num_examples=num_examples,
+            file_name_fn=lambda i: f"{i // 2:06d}_1{i % 2}.png",
+            num_examples=num_examples * 2,
             size=(3, 100, 200),
         )
         datasets_utils.create_image_folder(
             root=split_dir,
             name="image_3",
-            file_name_fn=lambda i: f"{i:06d}_10.png",
-            num_examples=num_examples,
+            file_name_fn=lambda i: f"{i // 2:06d}_1{i % 2}.png",
+            num_examples=num_examples * 2,
             size=(3, 100, 200),
         )
 
@@ -2762,7 +2762,7 @@ def inject_fake_data(self, tmpdir, config):
             datasets_utils.create_image_folder(
                 root=split_dir,
                 name="disp_occ_0",
-                file_name_fn=lambda i: f"{i:06d}.png",
+                file_name_fn=lambda i: f"{i:06d}_10.png",
                 num_examples=num_examples,
                 # Kitti2015 uses a single channel image for disparities
                 size=(1, 100, 200),
@@ -2771,7 +2771,7 @@ def inject_fake_data(self, tmpdir, config):
             datasets_utils.create_image_folder(
                 root=split_dir,
                 name="disp_occ_1",
-                file_name_fn=lambda i: f"{i:06d}.png",
+                file_name_fn=lambda i: f"{i:06d}_10.png",
                 num_examples=num_examples,
                 # Kitti2015 uses a single channel image for disparities
                 size=(1, 100, 200),
diff --git a/torchvision/datasets/_stereo_matching.py b/torchvision/datasets/_stereo_matching.py
index b07161d277c..feed695565c 100644
--- a/torchvision/datasets/_stereo_matching.py
+++ b/torchvision/datasets/_stereo_matching.py
@@ -334,13 +334,13 @@ def __init__(self, root: str, split: str = "train", transforms: Optional[Callabl
         verify_str_arg(split, "split", valid_values=("train", "test"))
 
         root = Path(root) / "Kitti2015" / (split + "ing")
-        left_img_pattern = str(root / "image_2" / "*.png")
-        right_img_pattern = str(root / "image_3" / "*.png")
+        left_img_pattern = str(root / "image_2" / "*_10.png")
+        right_img_pattern = str(root / "image_3" / "*_10.png")
         self._images = self._scan_pairs(left_img_pattern, right_img_pattern)
 
         if split == "train":
-            left_disparity_pattern = str(root / "disp_occ_0" / "*.png")
-            right_disparity_pattern = str(root / "disp_occ_1" / "*.png")
+            left_disparity_pattern = str(root / "disp_occ_0" / "*_10.png")
+            right_disparity_pattern = str(root / "disp_occ_1" / "*_10.png")
             self._disparities = self._scan_pairs(left_disparity_pattern, right_disparity_pattern)
         else:
             self._disparities = list((None, None) for _ in self._images)