- **Description**: COCO is a large-scale object detection, segmentation, and captioning dataset. > [!NOTE] > **Note:** - Some images from the train and validation sets don't have annotations. - Coco 2014 and 2017 uses the same images, but different train/val/test splits - The test split don't have any annotations (only images). - Coco defines 91 classes but the data only uses 80 classes. - Panotptic annotations defines defines 200 classes but only uses 133. - **Additional Documentation** : [Explore on Papers With Code](https://paperswithcode.com/dataset/coco) - **Homepage** : - **Source code** : [`tfds.object_detection.Coco`](https://github.com/tensorflow/datasets/tree/master/tensorflow_datasets/object_detection/coco.py) - **Versions**: - **`1.1.0`** (default): No release notes. - **Auto-cached** ([documentation](https://www.tensorflow.org/datasets/performances#auto-caching)): No - **Supervised keys** (See [`as_supervised` doc](https://www.tensorflow.org/datasets/api_docs/python/tfds/load#args)): `None` - **Citation**: @article{DBLP:journals/corr/LinMBHPRDZ14, author = {Tsung{-}Yi Lin and Michael Maire and Serge J. Belongie and Lubomir D. Bourdev and Ross B. Girshick and James Hays and Pietro Perona and Deva Ramanan and Piotr Doll{'{a} }r and C. Lawrence Zitnick}, title = {Microsoft {COCO:} Common Objects in Context}, journal = {CoRR}, volume = {abs/1405.0312}, year = {2014}, url = {http://arxiv.org/abs/1405.0312}, archivePrefix = {arXiv}, eprint = {1405.0312}, timestamp = {Mon, 13 Aug 2018 16:48:13 +0200}, biburl = {https://dblp.org/rec/bib/journals/corr/LinMBHPRDZ14}, bibsource = {dblp computer science bibliography, https://dblp.org} } ## coco/2014 (default config) - **Config description**: This version contains images, bounding boxes and labels for the 2014 version. - **Download size** : `37.57 GiB` - **Dataset size** : `37.35 GiB` - **Splits**: | Split | Examples | |---|---| | `'test'` | 40,775 | | `'test2015'` | 81,434 | | `'train'` | 82,783 | | `'validation'` | 40,504 | - **Feature structure**: FeaturesDict({ 'image': Image(shape=(None, None, 3), dtype=uint8), 'image/filename': Text(shape=(), dtype=string), 'image/id': int64, 'objects': Sequence({ 'area': int64, 'bbox': BBoxFeature(shape=(4,), dtype=float32), 'id': int64, 'is_crowd': bool, 'label': ClassLabel(shape=(), dtype=int64, num_classes=80), }), }) - **Feature documentation**: | Feature | Class | Shape | Dtype | Description | |---|---|---|---|---| | | FeaturesDict | | | | | image | Image | (None, None, 3) | uint8 | | | image/filename | Text | | string | | | image/id | Tensor | | int64 | | | objects | Sequence | | | | | objects/area | Tensor | | int64 | | | objects/bbox | BBoxFeature | (4,) | float32 | | | objects/id | Tensor | | int64 | | | objects/is_crowd | Tensor | | bool | | | objects/label | ClassLabel | | int64 | | - **Figure** ([tfds.show_examples](https://www.tensorflow.org/datasets/api_docs/python/tfds/visualization/show_examples)):  - **Examples** ([tfds.as_dataframe](https://www.tensorflow.org/datasets/api_docs/python/tfds/as_dataframe)): ## coco/2017 - **Config description**: This version contains images, bounding boxes and labels for the 2017 version. - **Download size** : `25.20 GiB` - **Dataset size** : `24.98 GiB` - **Splits**: | Split | Examples | |---|---| | `'test'` | 40,670 | | `'train'` | 118,287 | | `'validation'` | 5,000 | - **Feature structure**: FeaturesDict({ 'image': Image(shape=(None, None, 3), dtype=uint8), 'image/filename': Text(shape=(), dtype=string), 'image/id': int64, 'objects': Sequence({ 'area': int64, 'bbox': BBoxFeature(shape=(4,), dtype=float32), 'id': int64, 'is_crowd': bool, 'label': ClassLabel(shape=(), dtype=int64, num_classes=80), }), }) - **Feature documentation**: | Feature | Class | Shape | Dtype | Description | |---|---|---|---|---| | | FeaturesDict | | | | | image | Image | (None, None, 3) | uint8 | | | image/filename | Text | | string | | | image/id | Tensor | | int64 | | | objects | Sequence | | | | | objects/area | Tensor | | int64 | | | objects/bbox | BBoxFeature | (4,) | float32 | | | objects/id | Tensor | | int64 | | | objects/is_crowd | Tensor | | bool | | | objects/label | ClassLabel | | int64 | | - **Figure** ([tfds.show_examples](https://www.tensorflow.org/datasets/api_docs/python/tfds/visualization/show_examples)):  - **Examples** ([tfds.as_dataframe](https://www.tensorflow.org/datasets/api_docs/python/tfds/as_dataframe)): ## coco/2017_panoptic - **Config description**: This version contains images, bounding boxes and labels for the 2017 version. - **Download size** : `19.57 GiB` - **Dataset size** : `19.63 GiB` - **Splits**: | Split | Examples | |---|---| | `'train'` | 118,287 | | `'validation'` | 5,000 | - **Feature structure**: FeaturesDict({ 'image': Image(shape=(None, None, 3), dtype=uint8), 'image/filename': Text(shape=(), dtype=string), 'image/id': int64, 'panoptic_image': Image(shape=(None, None, 3), dtype=uint8), 'panoptic_image/filename': Text(shape=(), dtype=string), 'panoptic_objects': Sequence({ 'area': int64, 'bbox': BBoxFeature(shape=(4,), dtype=float32), 'id': int64, 'is_crowd': bool, 'label': ClassLabel(shape=(), dtype=int64, num_classes=133), }), }) - **Feature documentation**: | Feature | Class | Shape | Dtype | Description | |---|---|---|---|---| | | FeaturesDict | | | | | image | Image | (None, None, 3) | uint8 | | | image/filename | Text | | string | | | image/id | Tensor | | int64 | | | panoptic_image | Image | (None, None, 3) | uint8 | | | panoptic_image/filename | Text | | string | | | panoptic_objects | Sequence | | | | | panoptic_objects/area | Tensor | | int64 | | | panoptic_objects/bbox | BBoxFeature | (4,) | float32 | | | panoptic_objects/id | Tensor | | int64 | | | panoptic_objects/is_crowd | Tensor | | bool | | | panoptic_objects/label | ClassLabel | | int64 | | - **Figure** ([tfds.show_examples](https://www.tensorflow.org/datasets/api_docs/python/tfds/visualization/show_examples)): Not supported. - **Examples** ([tfds.as_dataframe](https://www.tensorflow.org/datasets/api_docs/python/tfds/as_dataframe)):