フリック
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。
論文より: 人気のハリウッド映画から 5003 画像データセットを自動的に収集しました。画像は、30 本の映画の 10 フレームごとに最先端の人物検出器を実行することによって取得されました。高い信頼度で検出された人物 (約 20,000 人の候補者) は、クラウドソーシング マーケットプレイスの Amazon Mechanical Turk に送信され、グラウンドトゥルースのラベルが取得されました。各画像には、5 人のトルコ人がそれぞれ 0.01 ドルで上半身の 10 個の関節にラベルを付ける注釈を付けました。外れ値の注釈に対して堅牢であるように、各画像で 5 の中央値のラベル付けが行われています。最後に、人物が遮蔽されているか、または著しく正面が欠けている場合、画像は手動で拒否されました。データの 20% (1016 枚の画像) をテスト用に確保しました。
スプリット | 例 |
---|
'test' | 1,016 |
'train' | 3,987 |
FeaturesDict({
'currframe': float64,
'image': Image(shape=(480, 720, 3), dtype=uint8),
'moviename': Text(shape=(), dtype=string),
'poselet_hit_idx': Sequence(uint16),
'torsobox': BBoxFeature(shape=(4,), dtype=float32),
'xcoords': Sequence(float64),
'ycoords': Sequence(float64),
})
特徴 | クラス | 形 | Dタイプ | 説明 |
---|
| 特徴辞書 | | | |
現在のフレーム | テンソル | | float64 | |
画像 | 画像 | (480、720、3) | uint8 | |
映画名 | 文章 | | 弦 | |
ポーズレット_ヒット_idx | シーケンス(テンソル) | (なし、) | uint16 | |
胴箱 | BBox機能 | (4,) | float32 | |
xcoords | シーケンス(テンソル) | (なし、) | float64 | |
ycoords | シーケンス(テンソル) | (なし、) | float64 | |
@inproceedings{modec13,
title={MODEC: Multimodal Decomposable Models for Human Pose Estimation},
author={Sapp, Benjamin and Taskar, Ben},
booktitle={In Proc. CVPR},
year={2013},
}
flic/small (デフォルト設定)

フリック/フル

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 2024-06-01 UTC。
[null,null,["最終更新日 2024-06-01 UTC。"],[],[],null,["# flic\n\n- **Description**:\n\nFrom the paper: We collected a 5003 image dataset automatically from popular\nHollywood movies. The images were obtained by running a state-of-the-art person\ndetector on every tenth frame of 30 movies. People detected with high confidence\n(roughly 20K candidates) were then sent to the crowdsourcing marketplace Amazon\nMechanical Turk to obtain groundtruthlabeling. Each image was annotated by five\nTurkers for $0.01 each to label 10 upperbody joints. The median-of-five labeling\nwas taken in each image to be robust to outlier annotation. Finally, images were\nrejected manually by us if the person was occluded or severely non-frontal. We\nset aside 20% (1016 images) of the data for testing.\n\n- **Additional Documentation** :\n [Explore on Papers With Code\n north_east](https://paperswithcode.com/dataset/flic)\n\n- **Homepage** :\n \u003chttps://bensapp.github.io/flic-dataset.html\u003e\n\n- **Source code** :\n [`tfds.datasets.flic.Builder`](https://github.com/tensorflow/datasets/tree/master/tensorflow_datasets/datasets/flic/flic_dataset_builder.py)\n\n- **Versions**:\n\n - **`2.0.0`** (default): No release notes.\n- **Dataset size** : `317.94 MiB`\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'` | 1,016 |\n| `'train'` | 3,987 |\n\n- **Feature structure**:\n\n FeaturesDict({\n 'currframe': float64,\n 'image': Image(shape=(480, 720, 3), dtype=uint8),\n 'moviename': Text(shape=(), dtype=string),\n 'poselet_hit_idx': Sequence(uint16),\n 'torsobox': BBoxFeature(shape=(4,), dtype=float32),\n 'xcoords': Sequence(float64),\n 'ycoords': Sequence(float64),\n })\n\n- **Feature documentation**:\n\n| Feature | Class | Shape | Dtype | Description |\n|-----------------|------------------|---------------|---------|-------------|\n| | FeaturesDict | | | |\n| currframe | Tensor | | float64 | |\n| image | Image | (480, 720, 3) | uint8 | |\n| moviename | Text | | string | |\n| poselet_hit_idx | Sequence(Tensor) | (None,) | uint16 | |\n| torsobox | BBoxFeature | (4,) | float32 | |\n| xcoords | Sequence(Tensor) | (None,) | float64 | |\n| ycoords | Sequence(Tensor) | (None,) | float64 | |\n\n- **Supervised keys** (See\n [`as_supervised` doc](https://www.tensorflow.org/datasets/api_docs/python/tfds/load#args)):\n `None`\n\n- **Citation**:\n\n @inproceedings{modec13,\n title={MODEC: Multimodal Decomposable Models for Human Pose Estimation},\n author={Sapp, Benjamin and Taskar, Ben},\n booktitle={In Proc. CVPR},\n year={2013},\n }\n\nflic/small (default config)\n---------------------------\n\n- **Config description**: Uses 5003 examples used in CVPR13 MODEC paper.\n\n- **Download size** : `286.35 MiB`\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\nflic/full\n---------\n\n- **Config description**: Uses 20928 examples, a superset of FLIC consisting\n of more difficult examples.\n\n- **Download size** : `1.10 GiB`\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..."]]