Skip to content

Latest commit

 

History

History
162 lines (127 loc) · 5.09 KB

emnist.md

File metadata and controls

162 lines (127 loc) · 5.09 KB
# `emnist`
  • Description:

The EMNIST dataset is a set of handwritten character digits derived from the NIST Special Database 19 and converted to a 28x28 pixel image format and dataset structure that directly matches the MNIST dataset.

Note: Like the original EMNIST data, images provided here are inverted horizontally and rotated 90 anti-clockwise. You can use tf.transpose within ds.map to convert the images to a human-friendlier format.

@article{cohen_afshar_tapson_schaik_2017,
    title={EMNIST: Extending MNIST to handwritten letters},
    DOI={10.1109/ijcnn.2017.7966217},
    journal={2017 International Joint Conference on Neural Networks (IJCNN)},
    author={Cohen, Gregory and Afshar, Saeed and Tapson, Jonathan and Schaik, Andre Van},
    year={2017}
}

emnist/byclass (default config)

  • Config description: EMNIST ByClass
  • Splits:
Split Examples
'test' 116,323
'train' 697,932
  • Features:
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=tf.uint8),
    'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=62),
})

emnist/bymerge

  • Config description: EMNIST ByMerge
  • Splits:
Split Examples
'test' 116,323
'train' 697,932
  • Features:
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=tf.uint8),
    'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=47),
})

emnist/balanced

  • Config description: EMNIST Balanced
  • Splits:
Split Examples
'test' 18,800
'train' 112,800
  • Features:
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=tf.uint8),
    'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=47),
})

emnist/letters

  • Config description: EMNIST Letters
  • Splits:
Split Examples
'test' 14,800
'train' 88,800
  • Features:
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=tf.uint8),
    'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=37),
})

emnist/digits

  • Config description: EMNIST Digits
  • Splits:
Split Examples
'test' 40,000
'train' 240,000
  • Features:
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=tf.uint8),
    'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=10),
})

emnist/mnist

  • Config description: EMNIST MNIST
  • Splits:
Split Examples
'test' 10,000
'train' 60,000
  • Features:
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=tf.uint8),
    'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=10),
})