kelapa

COCO adalah kumpulan data pendeteksian, segmentasi, dan keterangan objek berskala besar.

@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 (konfigurasi default)

  • Deskripsi konfigurasi : Versi ini berisi gambar, kotak pembatas, dan label untuk versi 2014.

  • Ukuran unduhan : 37.57 GiB

  • Ukuran dataset : 37.35 GiB

  • Perpecahan :

Membelah Contoh
'test' 40.775
'test2015' 81.434
'train' 82.783
'validation' 40.504
  • Struktur fitur :
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),
    }),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (Tidak ada, Tidak ada, 3) uint8
gambar/nama file Teks rangkaian
gambar/id Tensor int64
objek Urutan
benda/daerah Tensor int64
benda/bbox Fitur BBox (4,) float32
benda/id Tensor int64
objek/kerumunan_is Tensor bool
benda/label LabelKelas int64

Visualisasi

coco/2017

  • Deskripsi konfigurasi : Versi ini berisi gambar, kotak pembatas, dan label untuk versi 2017.

  • Ukuran unduhan : 25.20 GiB

  • Ukuran dataset : 24.98 GiB

  • Perpecahan :

Membelah Contoh
'test' 40.670
'train' 118.287
'validation' 5.000
  • Struktur fitur :
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),
    }),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (Tidak ada, Tidak ada, 3) uint8
gambar/nama file Teks rangkaian
gambar/id Tensor int64
objek Urutan
benda/daerah Tensor int64
benda/bbox Fitur BBox (4,) float32
benda/id Tensor int64
objek/kerumunan_is Tensor bool
benda/label LabelKelas int64

Visualisasi

coco/2017_panoptic

  • Deskripsi konfigurasi : Versi ini berisi gambar, kotak pembatas, dan label untuk versi 2017.

  • Ukuran unduhan : 19.57 GiB

  • Ukuran dataset : 19.63 GiB

  • Perpecahan :

Membelah Contoh
'train' 118.287
'validation' 5.000
  • Struktur fitur :
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),
    }),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (Tidak ada, Tidak ada, 3) uint8
gambar/nama file Teks rangkaian
gambar/id Tensor int64
panoptic_image Gambar (Tidak ada, Tidak ada, 3) uint8
panoptic_image/nama file Teks rangkaian
panoptic_objects Urutan
panoptic_objects/area Tensor int64
panoptic_objects/bbox Fitur BBox (4,) float32
panoptic_objects/id Tensor int64
panoptic_objects/is_crowd Tensor bool
panoptic_objects/label LabelKelas int64