visual_domain_decathlon

  • Description:

This contains the 10 datasets used in the Visual Domain Decathlon, part of the PASCAL in Detail Workshop Challenge (CVPR 2017). The goal of this challenge is to solve simultaneously ten image classification problems representative of very different visual domains.

Some of the datasets included here are also available as separate datasets in TFDS. However, notice that images were preprocessed for the Visual Domain Decathlon (resized isotropically to have a shorter size of 72 pixels) and might have different train/validation/test splits. Here we use the official splits for the competition.

@ONLINE{hakanbilensylvestrerebuffitomasjakab2017,
    author = "Hakan Bilen, Sylvestre Rebuffi, Tomas Jakab",
    title  = "Visual Domain Decathlon",
    year   = "2017",
    url    = "https://www.robots.ox.ac.uk/~vgg/decathlon/"
}

visual_domain_decathlon/aircraft (default config)

  • Config description: Data based on "Aircraft", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 20.96 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 3,333
'train' 3,334
'validation' 3,333
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=100),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/cifar100

  • Config description: Data based on "CIFAR-100", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 119.43 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 10,000
'train' 40,000
'validation' 10,000
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=100),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/daimlerpedcls

  • Config description: Data based on "Daimler Pedestrian Classification", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 68.35 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 19,600
'train' 23,520
'validation' 5,880
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=2),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/dtd

  • Config description: Data based on "Describable Textures", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 13.30 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 1,880
'train' 1,880
'validation' 1,880
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=47),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/gtsrb

  • Config description: Data based on "German Traffic Signs", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 80.58 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 12,630
'train' 31,367
'validation' 7,842
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=43),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/imagenet12

  • Config description: Data based on "Imagenet", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 6.11 GiB

  • Dataset size: 5.24 GiB

  • Auto-cached (documentation): No

  • Splits:

Split Examples
'test' 48,238
'train' 1,232,167
'validation' 49,000
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=1000),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/omniglot

  • Config description: Data based on "Omniglot", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 41.46 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 8,115
'train' 17,853
'validation' 6,492
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=1623),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/svhn

  • Config description: Data based on "Street View House Numbers", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 135.32 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 26,032
'train' 47,217
'validation' 26,040
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=10),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/ucf101

  • Config description: Data based on "UCF101 Dynamic Images", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 19.73 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 3,783
'train' 7,585
'validation' 1,952
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=101),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization

visual_domain_decathlon/vgg-flowers

  • Config description: Data based on "VGG-Flowers", with images resized isotropically to have a shorter size of 72 pixels.

  • Download size: 409.94 MiB

  • Dataset size: 20.87 MiB

  • Auto-cached (documentation): Yes

  • Splits:

Split Examples
'test' 6,149
'train' 1,020
'validation' 1,020
  • Feature structure:
FeaturesDict({
    'image': Image(shape=(None, None, 3), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=102),
    'name': Text(shape=(), dtype=string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
image Image (None, None, 3) uint8
label ClassLabel int64
name Text string

Visualization