SaulLu commited on
Commit
6aa1df0
1 Parent(s): 3f302b2

add configs

Browse files
Files changed (1) hide show
  1. Caltech-101.py +28 -2
Caltech-101.py CHANGED
@@ -156,6 +156,22 @@ class Caltech101(datasets.GeneratorBasedBuilder):
156
 
157
  VERSION = datasets.Version("1.0.0")
158
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
159
  def _info(self):
160
  return datasets.DatasetInfo(
161
  description=_DESCRIPTION,
@@ -188,6 +204,7 @@ class Caltech101(datasets.GeneratorBasedBuilder):
188
  gen_kwargs={
189
  "filepath": data_dir,
190
  "split": "train",
 
191
  },
192
  ),
193
  datasets.SplitGenerator(
@@ -195,13 +212,14 @@ class Caltech101(datasets.GeneratorBasedBuilder):
195
  gen_kwargs={
196
  "filepath": data_dir,
197
  "split": "test",
 
198
  },
199
  ),
200
  ]
201
 
202
- def _generate_examples(self, filepath, split):
203
  # Same stratagy as the one proposed in TF datasets: 30 random examples from each class are added to the train
204
- # split, and the remainder are added to the test split.
205
  # Source: https://github.com/tensorflow/datasets/blob/1106d587f97c4fca68c5b593dc7dc48c790ffa8c/tensorflow_datasets/image_classification/caltech.py#L88-L140
206
 
207
  is_train_split = split == "train"
@@ -212,6 +230,7 @@ class Caltech101(datasets.GeneratorBasedBuilder):
212
  np.random.seed(1234)
213
 
214
  for class_dir in data_dir.iterdir():
 
215
  fnames = [
216
  image_path
217
  for image_path in class_dir.iterdir()
@@ -231,6 +250,13 @@ class Caltech101(datasets.GeneratorBasedBuilder):
231
  test_fnames = set(fnames).difference(train_fnames)
232
  fnames_to_emit = train_fnames if is_train_split else test_fnames
233
 
 
 
 
 
 
 
 
234
  for image_file in fnames_to_emit:
235
  record = {
236
  "image": str(image_file),
 
156
 
157
  VERSION = datasets.Version("1.0.0")
158
 
159
+ _BUILDER_CONFIG_WITH_BACKGROUND = datasets.BuilderConfig(
160
+ name="with_background_category",
161
+ version=VERSION,
162
+ description="Dataset containing only the 101 categories.",
163
+ )
164
+ _BUILDER_CONFIG_WITHOUT_BACKGROUND = datasets.BuilderConfig(
165
+ name="without_background_category",
166
+ version=VERSION,
167
+ description="Dataset containing the 101 categories and the additonnal background one.",
168
+ )
169
+
170
+ BUILDER_CONFIGS = [
171
+ _BUILDER_CONFIG_WITH_BACKGROUND,
172
+ _BUILDER_CONFIG_WITHOUT_BACKGROUND,
173
+ ]
174
+
175
  def _info(self):
176
  return datasets.DatasetInfo(
177
  description=_DESCRIPTION,
 
204
  gen_kwargs={
205
  "filepath": data_dir,
206
  "split": "train",
207
+ "config_name": self.config.name,
208
  },
209
  ),
210
  datasets.SplitGenerator(
 
212
  gen_kwargs={
213
  "filepath": data_dir,
214
  "split": "test",
215
+ "config_name": self.config.name,
216
  },
217
  ),
218
  ]
219
 
220
+ def _generate_examples(self, filepath, split, config_name):
221
  # Same stratagy as the one proposed in TF datasets: 30 random examples from each class are added to the train
222
+ # split, and the remainder are added to the test split.
223
  # Source: https://github.com/tensorflow/datasets/blob/1106d587f97c4fca68c5b593dc7dc48c790ffa8c/tensorflow_datasets/image_classification/caltech.py#L88-L140
224
 
225
  is_train_split = split == "train"
 
230
  np.random.seed(1234)
231
 
232
  for class_dir in data_dir.iterdir():
233
+ # print(class_dir)
234
  fnames = [
235
  image_path
236
  for image_path in class_dir.iterdir()
 
250
  test_fnames = set(fnames).difference(train_fnames)
251
  fnames_to_emit = train_fnames if is_train_split else test_fnames
252
 
253
+ if (
254
+ class_dir.name == "BACKGROUND_Google"
255
+ and config_name == self._BUILDER_CONFIG_WITHOUT_BACKGROUND.name
256
+ ):
257
+ print("skip BACKGROUND_Google")
258
+ continue
259
+
260
  for image_file in fnames_to_emit:
261
  record = {
262
  "image": str(image_file),