ウェイクビジョン
コレクションでコンテンツを整理
必要に応じて、コンテンツの保存と分類を行います。
注: このデータセットは最近追加されたもので、 tfds-nightly
パッケージnights_stay でのみ使用できます。 Wake Vision は、600 万を超える画像を特徴とする大規模で高品質のデータセットであり、現在の tinyML データセット (100 倍) の規模と多様性を大幅に超えています。このデータセットには、各画像に人物が含まれているかどうかの注釈が付いた画像が含まれています。さらに、公平性と堅牢性を評価するための包括的なきめの細かいベンチマークが組み込まれており、知覚される性別、知覚される年齢、被写体の距離、照明条件、描写をカバーします。 Wake Vision ラベルは、CC BY 4.0 ライセンスに基づいて Google LLC からライセンスを取得した Open Image の注釈から派生しています。画像は CC BY 2.0 ライセンスを持つものとしてリストされています。 Open Images からの注記: 「クリエイティブ コモンズ表示ライセンスに基づいてライセンスが付与されている画像を特定しようとしましたが、各画像のライセンス ステータスに関していかなる表明も保証も行いません。各画像のライセンスはご自身で確認してください。」
スプリット 例 'test'
55,763 'train_large'
5,760,428 'train_quality'
1,248,230 'validation'
18,582
FeaturesDict ({
'age_unknown' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'body_part' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'bright' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'dark' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'depiction' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'far' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'filename' : Text ( shape = (), dtype = string ),
'gender_unknown' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'image' : Image ( shape = ( None , None , 3 ), dtype = uint8 ),
'medium_distance' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'middle_age' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'near' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'non-person_depiction' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'non-person_non-depiction' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'normal_lighting' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'older' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'person' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'person_depiction' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'predominantly_female' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'predominantly_male' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
'young' : ClassLabel ( shape = (), dtype = int64 , num_classes = 2 ),
})
特徴 クラス 形 Dタイプ 説明 特徴辞書 年齢不明 クラスラベル int64 ボディパーツ クラスラベル int64 明るい クラスラベル int64 暗い クラスラベル int64 描写 クラスラベル int64 遠い クラスラベル int64 ファイル名 文章 弦 性別不明 クラスラベル int64 画像 画像 (なし、なし、3) uint8 中距離 クラスラベル int64 中年 クラスラベル int64 近く クラスラベル int64 人物以外の描写 クラスラベル int64 非人物_非描写 クラスラベル int64 通常_照明 クラスラベル int64 年上の クラスラベル int64 人 クラスラベル int64 人物の描写 クラスラベル int64 主に女性 クラスラベル int64 主に男性 クラスラベル int64 若い クラスラベル int64
@article { banbury2024wake ,
title = { Wake Vision : A Large - scale , Diverse Dataset and Benchmark Suite for TinyML Person Detection } ,
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 } ,
journal = { arXiv preprint arXiv : 2405.00892 } ,
year = { 2024 }
}
特に記載のない限り、このページのコンテンツはクリエイティブ・コモンズの表示 4.0 ライセンス により使用許諾されます。コードサンプルは Apache 2.0 ライセンス により使用許諾されます。詳しくは、Google Developers サイトのポリシー をご覧ください。Java は Oracle および関連会社の登録商標です。
最終更新日 2024-12-14 UTC。
[null,null,["最終更新日 2024-12-14 UTC。"],[],[],null,["# wake_vision\n\n\u003cbr /\u003e\n\n- **Description**:\n\nWake Vision is a large, high-quality dataset featuring over 6 million images,\nsignificantly exceeding the scale and diversity of current tinyML datasets\n(100x). This dataset includes images with annotations of whether each image\ncontains a person. Additionally, it incorporates a comprehensive fine-grained\nbenchmark to assess fairness and robustness, covering perceived gender,\nperceived age, subject distance, lighting conditions, and depictions. The Wake\nVision labels are derived from Open Image's annotations which are licensed by\nGoogle LLC under CC BY 4.0 license. The images are listed as having a CC BY 2.0\nlicense. Note from Open Images: \"while we tried to identify images that are\nlicensed under a Creative Commons Attribution license, we make no\nrepresentations or warranties regarding the license status of each image and you\nshould verify the license for each image yourself.\"\n\n- **Homepage** :\n \u003chttps://dataverse.harvard.edu/dataset.xhtml?persistentId=doi%3A10.7910%2FDVN%2F1HOPXC\u003e\n\n- **Source code** :\n [`tfds.datasets.wake_vision.Builder`](https://github.com/tensorflow/datasets/tree/master/tensorflow_datasets/datasets/wake_vision/wake_vision_dataset_builder.py)\n\n- **Versions**:\n\n - **`1.0.0`** (default): Initial TensorFlow Datasets release. Note that this is based on the 2.0 version of Wake Vision on Harvard Dataverse.\n- **Download size** : `Unknown size`\n\n- **Dataset size** : `239.25 GiB`\n\n- **Auto-cached**\n ([documentation](https://www.tensorflow.org/datasets/performances#auto-caching)):\n No\n\n- **Splits**:\n\n| Split | Examples |\n|-------------------|-----------|\n| `'test'` | 55,763 |\n| `'train_large'` | 5,760,428 |\n| `'train_quality'` | 1,248,230 |\n| `'validation'` | 18,582 |\n\n- **Feature structure**:\n\n FeaturesDict({\n 'age_unknown': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'body_part': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'bright': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'dark': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'depiction': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'far': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'filename': Text(shape=(), dtype=string),\n 'gender_unknown': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'image': Image(shape=(None, None, 3), dtype=uint8),\n 'medium_distance': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'middle_age': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'near': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'non-person_depiction': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'non-person_non-depiction': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'normal_lighting': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'older': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'person': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'person_depiction': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'predominantly_female': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'predominantly_male': ClassLabel(shape=(), dtype=int64, num_classes=2),\n 'young': ClassLabel(shape=(), dtype=int64, num_classes=2),\n })\n\n- **Feature documentation**:\n\n| Feature | Class | Shape | Dtype | Description |\n|--------------------------|--------------|-----------------|--------|-------------|\n| | FeaturesDict | | | |\n| age_unknown | ClassLabel | | int64 | |\n| body_part | ClassLabel | | int64 | |\n| bright | ClassLabel | | int64 | |\n| dark | ClassLabel | | int64 | |\n| depiction | ClassLabel | | int64 | |\n| far | ClassLabel | | int64 | |\n| filename | Text | | string | |\n| gender_unknown | ClassLabel | | int64 | |\n| image | Image | (None, None, 3) | uint8 | |\n| medium_distance | ClassLabel | | int64 | |\n| middle_age | ClassLabel | | int64 | |\n| near | ClassLabel | | int64 | |\n| non-person_depiction | ClassLabel | | int64 | |\n| non-person_non-depiction | ClassLabel | | int64 | |\n| normal_lighting | ClassLabel | | int64 | |\n| older | ClassLabel | | int64 | |\n| person | ClassLabel | | int64 | |\n| person_depiction | ClassLabel | | int64 | |\n| predominantly_female | ClassLabel | | int64 | |\n| predominantly_male | ClassLabel | | int64 | |\n| young | ClassLabel | | int64 | |\n\n- **Supervised keys** (See\n [`as_supervised` doc](https://www.tensorflow.org/datasets/api_docs/python/tfds/load#args)):\n `('image', 'person')`\n\n- **Figure**\n ([tfds.show_examples](https://www.tensorflow.org/datasets/api_docs/python/tfds/visualization/show_examples)):\n\n- **Examples** ([tfds.as_dataframe](https://www.tensorflow.org/datasets/api_docs/python/tfds/as_dataframe)):\n\nDisplay examples... \n\n- **Citation**:\n\n @article{banbury2024wake,\n title={Wake Vision: A Large-scale, Diverse Dataset and Benchmark Suite for TinyML Person Detection},\n 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},\n journal={arXiv preprint arXiv:2405.00892},\n year={2024}\n }"]]