fix

README.md

@@ -45,8 +45,10 @@ dataset_info:
         dtype: string
   splits:
   - name: train
-  download_size: 313332745
-  dataset_size: 0
+    num_bytes: 34542
+    num_examples: 52
+  download_size: 313334253
+  dataset_size: 34542
 - config_name: video_02
   features:
   - name: id
@@ -83,8 +85,10 @@ dataset_info:
         dtype: string
   splits:
   - name: train
-  download_size: 67366481
-  dataset_size: 0
+    num_bytes: 38528
+    num_examples: 58
+  download_size: 67368145
+  dataset_size: 38528
 - config_name: video_03
   features:
   - name: id
@@ -121,8 +125,10 @@ dataset_info:
         dtype: string
   splits:
   - name: train
-  download_size: 148416577
-  dataset_size: 0
+    num_bytes: 32550
+    num_examples: 49
+  download_size: 148418007
+  dataset_size: 32550
 ---
 
 # Dogs Video Object Tracking Dataset

dogs-video-object-tracking-dataset.py

@@ -80,16 +80,13 @@ class DogsVideoObjectTrackingDataset(datasets.GeneratorBasedBuilder):
 
     def _split_generators(self, dl_manager):
         images = dl_manager.download_and_extract(f"{_DATA}{self.config.name}.zip")
-        masks = dl_manager.download_and_extract(f"{_DATA}{self.config.name}_masks.zip")
         annotations = dl_manager.download(f"{_DATA}{self.config.name}.xml")
         images = dl_manager.iter_files(images)
-        masks = dl_manager.iter_files(masks)
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
                     "images": images,
-                    "masks": masks,
                     "annotations": annotations,
                 },
             ),
@@ -155,11 +152,11 @@ class DogsVideoObjectTrackingDataset(datasets.GeneratorBasedBuilder):
 
         return shape_data
 
-    def _generate_examples(self, images, masks, annotations):
+    def _generate_examples(self, images, annotations):
         tree = ET.parse(annotations)
         root = tree.getroot()
 
-        for idx, (image, mask) in enumerate(zip(images, masks)):
+        for idx, image in enumerate(images):
             if "images" in image:
                 i = image.split("_")[-1][:2]
                 image_name = image.split("/")[3]
@@ -172,7 +169,7 @@ class DogsVideoObjectTrackingDataset(datasets.GeneratorBasedBuilder):
                 yield idx, {
                     "id": image_id,
                     "name": image,
-                    "image": f"{images}/image",
-                    "mask": f"{masks}/mask",
+                    "image": f"{images}/images/{image}",
+                    "mask": f"{images}/masks/{image}",
                     "shapes": shapes,
                 }