Skip to content

Commit ff7171e

Browse files
author
The TensorFlow Datasets Authors
committed
Automated documentation update.
PiperOrigin-RevId: 663031637
1 parent 3b670a1 commit ff7171e

File tree

3 files changed

+175
-0
lines changed

3 files changed

+175
-0
lines changed

‎docs/catalog/_toc.yaml

Lines changed: 23 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -36,6 +36,9 @@ toc:
3636
- section:
3737
- path: /datasets/catalog/dices
3838
title: dices
39+
- path: /datasets/catalog/wake_vision
40+
status: nightly
41+
title: wake_vision
3942
title: Age
4043
- section:
4144
- path: /datasets/catalog/ag_news_subset
@@ -104,6 +107,9 @@ toc:
104107
title: dices
105108
- path: /datasets/catalog/sift1m
106109
title: sift1m
110+
- path: /datasets/catalog/wake_vision
111+
status: nightly
112+
title: wake_vision
107113
title: Categorical
108114
- section:
109115
- path: /datasets/catalog/ai2_arc_with_ir
@@ -189,6 +195,11 @@ toc:
189195
- path: /datasets/catalog/scientific_papers
190196
title: scientific_papers
191197
title: Document summarization
198+
- section:
199+
- path: /datasets/catalog/wake_vision
200+
status: nightly
201+
title: wake_vision
202+
title: Facial attributes
192203
- section:
193204
- path: /datasets/catalog/caltech101
194205
title: caltech101
@@ -202,10 +213,16 @@ toc:
202213
title: stl10
203214
- path: /datasets/catalog/sun397
204215
title: sun397
216+
- path: /datasets/catalog/wake_vision
217+
status: nightly
218+
title: wake_vision
205219
title: Fine grained image classification
206220
- section:
207221
- path: /datasets/catalog/dices
208222
title: dices
223+
- path: /datasets/catalog/wake_vision
224+
status: nightly
225+
title: wake_vision
209226
title: Gender
210227
- section:
211228
- path: /datasets/catalog/ogbg_molpcba
@@ -366,6 +383,9 @@ toc:
366383
title: tf_flowers
367384
- path: /datasets/catalog/the300w_lp
368385
title: the300w_lp
386+
- path: /datasets/catalog/wake_vision
387+
status: nightly
388+
title: wake_vision
369389
title: Image
370390
- section:
371391
- path: /datasets/catalog/abstract_reasoning
@@ -500,6 +520,9 @@ toc:
500520
title: uc_merced
501521
- path: /datasets/catalog/visual_domain_decathlon
502522
title: visual_domain_decathlon
523+
- path: /datasets/catalog/wake_vision
524+
status: nightly
525+
title: wake_vision
503526
title: Image classification
504527
- section:
505528
- path: /datasets/catalog/imagenet2012

‎docs/catalog/overview.md

Lines changed: 17 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -49,6 +49,8 @@ for ex in tfds.load('cifar10', split='train'):
4949
### `Age`
5050

5151
* [`dices`](dices.md)
52+
* [`wake_vision`](wake_vision.md)
53+
<span class="material-icons" title="Available only in the tfds-nightly package">nights_stay</span>
5254

5355
### `Anomaly detection`
5456

@@ -91,6 +93,8 @@ for ex in tfds.load('cifar10', split='train'):
9193

9294
* [`dices`](dices.md)
9395
* [`sift1m`](sift1m.md)
96+
* [`wake_vision`](wake_vision.md)
97+
<span class="material-icons" title="Available only in the tfds-nightly package">nights_stay</span>
9498

9599
### `Common sense reasoning`
96100

@@ -154,6 +158,11 @@ for ex in tfds.load('cifar10', split='train'):
154158
* [`newsroom`](newsroom.md)
155159
* [`scientific_papers`](scientific_papers.md)
156160

161+
### `Facial attributes`
162+
163+
* [`wake_vision`](wake_vision.md)
164+
<span class="material-icons" title="Available only in the tfds-nightly package">nights_stay</span>
165+
157166
### `Fine grained image classification`
158167

159168
* [`caltech101`](caltech101.md)
@@ -162,10 +171,14 @@ for ex in tfds.load('cifar10', split='train'):
162171
* [`stanford_dogs`](stanford_dogs.md)
163172
* [`stl10`](stl10.md)
164173
* [`sun397`](sun397.md)
174+
* [`wake_vision`](wake_vision.md)
175+
<span class="material-icons" title="Available only in the tfds-nightly package">nights_stay</span>
165176

166177
### `Gender`
167178

168179
* [`dices`](dices.md)
180+
* [`wake_vision`](wake_vision.md)
181+
<span class="material-icons" title="Available only in the tfds-nightly package">nights_stay</span>
169182

170183
### `Graph`
171184

@@ -252,6 +265,8 @@ for ex in tfds.load('cifar10', split='train'):
252265
* [`symmetric_solids`](symmetric_solids.md)
253266
* [`tf_flowers`](tf_flowers.md)
254267
* [`the300w_lp`](the300w_lp.md)
268+
* [`wake_vision`](wake_vision.md)
269+
<span class="material-icons" title="Available only in the tfds-nightly package">nights_stay</span>
255270

256271
### `Image classification`
257272

@@ -321,6 +336,8 @@ for ex in tfds.load('cifar10', split='train'):
321336
* [`svhn_cropped`](svhn_cropped.md)
322337
* [`uc_merced`](uc_merced.md)
323338
* [`visual_domain_decathlon`](visual_domain_decathlon.md)
339+
* [`wake_vision`](wake_vision.md)
340+
<span class="material-icons" title="Available only in the tfds-nightly package">nights_stay</span>
324341

325342
### `Image clustering`
326343

‎docs/catalog/wake_vision.md

Lines changed: 135 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,135 @@
1+
<div itemscope itemtype="http://schema.org/Dataset">
2+
<div itemscope itemprop="includedInDataCatalog" itemtype="http://schema.org/DataCatalog">
3+
<meta itemprop="name" content="TensorFlow Datasets" />
4+
</div>
5+
<meta itemprop="name" content="wake_vision" />
6+
<meta itemprop="description" content="Wake Vision is a large, high-quality dataset featuring over 6 million images,&#10;significantly exceeding the scale and diversity of current tinyML datasets&#10;(100x). This dataset includes images with annotations of whether each image&#10;contains a person. Additionally, it incorporates a comprehensive fine-grained&#10;benchmark to assess fairness and robustness, covering perceived gender,&#10;perceived age, subject distance, lighting conditions, and depictions. The Wake&#10;Vision labels are derived from Open Image&#x27;s annotations which are licensed by&#10;Google LLC under CC BY 4.0 license. The images are listed as having a CC BY 2.0&#10;license. Note from Open Images: &quot;while we tried to identify images that are&#10;licensed under a Creative Commons Attribution license, we make no&#10;representations or warranties regarding the license status of each image and you&#10;should verify the license for each image yourself.&quot;&#10;&#10;To use this dataset:&#10;&#10;```python&#10;import tensorflow_datasets as tfds&#10;&#10;ds = tfds.load(&#x27;wake_vision&#x27;, split=&#x27;train&#x27;)&#10;for ex in ds.take(4):&#10; print(ex)&#10;```&#10;&#10;See [the guide](https://www.tensorflow.org/datasets/overview) for more&#10;informations on [tensorflow_datasets](https://www.tensorflow.org/datasets).&#10;&#10;" />
7+
<meta itemprop="url" content="https://www.tensorflow.org/datasets/catalog/wake_vision" />
8+
<meta itemprop="sameAs" content="https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi%3A10.7910%2FDVN%2F1HOPXC" />
9+
<meta itemprop="citation" content="@article{banbury2024wake,&#10; title={Wake Vision: A Large-scale, Diverse Dataset and Benchmark Suite for TinyML Person Detection},&#10; author={Banbury, Colby and Njor, Emil and Stewart, Matthew and Warden, Pete and Kudlur, Manjunath and Jeffries, Nat and Fafoutis, Xenofon and Reddi, Vijay Janapa},&#10; journal={arXiv preprint arXiv:2405.00892},&#10; year={2024}&#10;}" />
10+
</div>
11+
12+
# `wake_vision`
13+
14+
15+
Note: This dataset was added recently and is only available in our
16+
`tfds-nightly` package
17+
<span class="material-icons" title="Available only in the tfds-nightly package">nights_stay</span>.
18+
19+
* **Description**:
20+
21+
Wake Vision is a large, high-quality dataset featuring over 6 million images,
22+
significantly exceeding the scale and diversity of current tinyML datasets
23+
(100x). This dataset includes images with annotations of whether each image
24+
contains a person. Additionally, it incorporates a comprehensive fine-grained
25+
benchmark to assess fairness and robustness, covering perceived gender,
26+
perceived age, subject distance, lighting conditions, and depictions. The Wake
27+
Vision labels are derived from Open Image's annotations which are licensed by
28+
Google LLC under CC BY 4.0 license. The images are listed as having a CC BY 2.0
29+
license. Note from Open Images: "while we tried to identify images that are
30+
licensed under a Creative Commons Attribution license, we make no
31+
representations or warranties regarding the license status of each image and you
32+
should verify the license for each image yourself."
33+
34+
* **Homepage**:
35+
[https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi%3A10.7910%2FDVN%2F1HOPXC](https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi%3A10.7910%2FDVN%2F1HOPXC)
36+
37+
* **Source code**:
38+
[`tfds.datasets.wake_vision.Builder`](https://github.com/tensorflow/datasets/tree/master/tensorflow_datasets/datasets/wake_vision/wake_vision_dataset_builder.py)
39+
40+
* **Versions**:
41+
42+
* **`1.0.0`** (default): Initial TensorFlow Datasets release. Note that
43+
this is based on the 2.0 version of Wake Vision on Harvard Dataverse.
44+
45+
* **Download size**: `Unknown size`
46+
47+
* **Dataset size**: `Unknown size`
48+
49+
* **Auto-cached**
50+
([documentation](https://www.tensorflow.org/datasets/performances#auto-caching)):
51+
Unknown
52+
53+
* **Splits**:
54+
55+
Split | Examples
56+
:---- | -------:
57+
58+
* **Feature structure**:
59+
60+
```python
61+
FeaturesDict({
62+
'age_unknown': ClassLabel(shape=(), dtype=int64, num_classes=2),
63+
'body_part': ClassLabel(shape=(), dtype=int64, num_classes=2),
64+
'bright': ClassLabel(shape=(), dtype=int64, num_classes=2),
65+
'dark': ClassLabel(shape=(), dtype=int64, num_classes=2),
66+
'depiction': ClassLabel(shape=(), dtype=int64, num_classes=2),
67+
'far': ClassLabel(shape=(), dtype=int64, num_classes=2),
68+
'filename': Text(shape=(), dtype=string),
69+
'gender_unknown': ClassLabel(shape=(), dtype=int64, num_classes=2),
70+
'image': Image(shape=(None, None, 3), dtype=uint8),
71+
'medium_distance': ClassLabel(shape=(), dtype=int64, num_classes=2),
72+
'middle_age': ClassLabel(shape=(), dtype=int64, num_classes=2),
73+
'near': ClassLabel(shape=(), dtype=int64, num_classes=2),
74+
'non-person_depiction': ClassLabel(shape=(), dtype=int64, num_classes=2),
75+
'non-person_non-depiction': ClassLabel(shape=(), dtype=int64, num_classes=2),
76+
'normal_lighting': ClassLabel(shape=(), dtype=int64, num_classes=2),
77+
'older': ClassLabel(shape=(), dtype=int64, num_classes=2),
78+
'person': ClassLabel(shape=(), dtype=int64, num_classes=2),
79+
'person_depiction': ClassLabel(shape=(), dtype=int64, num_classes=2),
80+
'predominantly_female': ClassLabel(shape=(), dtype=int64, num_classes=2),
81+
'predominantly_male': ClassLabel(shape=(), dtype=int64, num_classes=2),
82+
'young': ClassLabel(shape=(), dtype=int64, num_classes=2),
83+
})
84+
```
85+
86+
* **Feature documentation**:
87+
88+
Feature | Class | Shape | Dtype | Description
89+
:----------------------- | :----------- | :-------------- | :----- | :----------
90+
| FeaturesDict | | |
91+
age_unknown | ClassLabel | | int64 |
92+
body_part | ClassLabel | | int64 |
93+
bright | ClassLabel | | int64 |
94+
dark | ClassLabel | | int64 |
95+
depiction | ClassLabel | | int64 |
96+
far | ClassLabel | | int64 |
97+
filename | Text | | string |
98+
gender_unknown | ClassLabel | | int64 |
99+
image | Image | (None, None, 3) | uint8 |
100+
medium_distance | ClassLabel | | int64 |
101+
middle_age | ClassLabel | | int64 |
102+
near | ClassLabel | | int64 |
103+
non-person_depiction | ClassLabel | | int64 |
104+
non-person_non-depiction | ClassLabel | | int64 |
105+
normal_lighting | ClassLabel | | int64 |
106+
older | ClassLabel | | int64 |
107+
person | ClassLabel | | int64 |
108+
person_depiction | ClassLabel | | int64 |
109+
predominantly_female | ClassLabel | | int64 |
110+
predominantly_male | ClassLabel | | int64 |
111+
young | ClassLabel | | int64 |
112+
113+
* **Supervised keys** (See
114+
[`as_supervised` doc](https://www.tensorflow.org/datasets/api_docs/python/tfds/load#args)):
115+
`('image', 'person')`
116+
117+
* **Figure**
118+
([tfds.show_examples](https://www.tensorflow.org/datasets/api_docs/python/tfds/visualization/show_examples)):
119+
Not supported.
120+
121+
* **Examples**
122+
([tfds.as_dataframe](https://www.tensorflow.org/datasets/api_docs/python/tfds/as_dataframe)):
123+
Missing.
124+
125+
* **Citation**:
126+
127+
```
128+
@article{banbury2024wake,
129+
title={Wake Vision: A Large-scale, Diverse Dataset and Benchmark Suite for TinyML Person Detection},
130+
author={Banbury, Colby and Njor, Emil and Stewart, Matthew and Warden, Pete and Kudlur, Manjunath and Jeffries, Nat and Fafoutis, Xenofon and Reddi, Vijay Janapa},
131+
journal={arXiv preprint arXiv:2405.00892},
132+
year={2024}
133+
}
134+
```
135+

0 commit comments

Comments
 (0)