Add "display_image" feature to robotics dataset importer builder.

PiperOrigin-RevId: 638110575

tensorflow_datasets/robotics/dataset_importer_builder.py

@@ -54,6 +54,8 @@ class DatasetImporterBuilder(
       'ssot_session_key',
   ]
 
+  images_from_observation_dict = {}
+
 
   @abc.abstractmethod
   def get_description(self):
@@ -83,9 +85,15 @@ def _info(self) -> tfds.core.DatasetInfo:
 
     tmp = dict(features)
 
+    # add all image features from observations to a new featuresdict
+    self.images_from_observation_dict = self.get_images_from_observation_dict()
+    if self.images_from_observation_dict:
+      tmp['display_image'] = self.images_from_observation_dict
+
     for key in self.KEYS_TO_STRIP:
       if key in tmp:
         del tmp[key]
+
     features = tfds.features.FeaturesDict(tmp)
 
     return tfds.core.DatasetInfo(
@@ -120,15 +128,28 @@ def _generate_examples(
     def converter_fn(example):
       # Decode the RLDS Episode and transform it to numpy.
       example_out = dict(example)
+
       example_out['steps'] = tf.data.Dataset.from_tensor_slices(
           example_out['steps']
       ).map(decode_fn)
+
       steps = list(iter(example_out['steps'].take(-1)))
       example_out['steps'] = steps
 
       example_out = dataset_utils.as_numpy(example_out)
+      first_step = example_out['steps'][0]
+      image_feature_dict = {}
+
+      for feature_name in self.images_from_observation_dict:
+        image_feature_dict[feature_name] = first_step['observation'][
+            feature_name
+        ]
+
+      if image_feature_dict:
+        example_out['display_image'] = image_feature_dict
 
       example_id = example_out['tfds_id'].decode('utf-8')
+
       del example_out['tfds_id']
       for key in self.KEYS_TO_STRIP:
         if key in example_out:
@@ -148,3 +169,17 @@ def get_ds_builder(self):
     ds_location = self.get_dataset_location()
     ds_builder = tfds.builder_from_directory(ds_location)
     return ds_builder
+
+  def get_images_from_observation_dict(self):
+    features = self.get_ds_builder().info.features
+    tmp = dict(features)
+    images_from_observation = {}
+    if 'steps' in tmp and 'observation' in tmp['steps']:
+      observation = tmp['steps']['observation']
+      for feature_name, feature_data in observation.items():
+        if isinstance(feature_data, tfds.features.Image):
+          images_from_observation[feature_name] = feature_data
+      images_from_observation_dict = tfds.features.FeaturesDict(
+          images_from_observation
+      )
+      return images_from_observation_dict