coco

@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)

  • Description:

COCO is a large-scale object detection, segmentation, and captioning dataset. This version contains images, bounding boxes " and labels for the 2014 version. 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.

  • Config description: COCO is a large-scale object detection, segmentation, and captioning dataset. This version contains images, bounding boxes " and labels for the 2014 version. 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.

  • Download size: 37.57 GiB

  • Splits:

Split Examples
'test' 40,775
'test2015' 81,434
'train' 82,783
'validation' 40,504
  • Features:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=tf.uint8),
    'image/filename': Text(shape=(), dtype=tf.string),
    'image/id': tf.int64,
    'objects': Sequence({
        'area': tf.int64,
        'bbox': BBoxFeature(shape=(4,), dtype=tf.float32),
        'id': tf.int64,
        'is_crowd': tf.bool,
        'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=80),
    }),
})

Visualization

coco/2017

  • Description:

COCO is a large-scale object detection, segmentation, and captioning dataset. This version contains images, bounding boxes " and labels for the 2017 version. 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.

  • Config description: COCO is a large-scale object detection, segmentation, and captioning dataset. This version contains images, bounding boxes " and labels for the 2017 version. 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.

  • Download size: 25.20 GiB

  • Splits:

Split Examples
'test' 40,670
'train' 118,287
'validation' 5,000
  • Features:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=tf.uint8),
    'image/filename': Text(shape=(), dtype=tf.string),
    'image/id': tf.int64,
    'objects': Sequence({
        'area': tf.int64,
        'bbox': BBoxFeature(shape=(4,), dtype=tf.float32),
        'id': tf.int64,
        'is_crowd': tf.bool,
        'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=80),
    }),
})

Visualization

coco/2017_panoptic

  • Description:

COCO is a large-scale object detection, segmentation, and captioning dataset. This version contains images, bounding boxes " and labels for the 2017 version. 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.

  • Config description: COCO is a large-scale object detection, segmentation, and captioning dataset. This version contains images, bounding boxes " and labels for the 2017 version. 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.

  • Download size: 19.57 GiB

  • Splits:

Split Examples
'train' 118,287
'validation' 5,000
  • Features:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=tf.uint8),
    'image/filename': Text(shape=(), dtype=tf.string),
    'image/id': tf.int64,
    'panoptic_image': Image(shape=(None, None, 3), dtype=tf.uint8),
    'panoptic_image/filename': Text(shape=(), dtype=tf.string),
    'panoptic_objects': Sequence({
        'area': tf.int64,
        'bbox': BBoxFeature(shape=(4,), dtype=tf.float32),
        'id': tf.int64,
        'is_crowd': tf.bool,
        'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=133),
    }),
})