fix: load script

dogs-video-object-tracking-dataset.py

@@ -154,7 +154,8 @@ class DogsVideoObjectTrackingDataset(datasets.GeneratorBasedBuilder):
         root = tree.getroot()
 
         for idx, file in enumerate(sorted(os.listdir(f"{data}/images"))):
-            img = self.extract_shapes_from_tracks(root, file, idx)
+            i = file.split(".")[0].split("_")[-1]
+            img = self.extract_shapes_from_tracks(root, file, i)
 
             image_id = img.get("id")
             name = img.get("name")