emnist

Dataset EMNIST adalah sekumpulan digit karakter tulisan tangan yang berasal dari NIST Special Database 19 dan dikonversi ke format gambar 28x28 piksel dan struktur dataset yang secara langsung cocok dengan dataset MNIST.

@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 (konfigurasi default)

  • Deskripsi konfigurasi : EMNIST ByClass

  • Ukuran dataset : 349.16 MiB

  • Di-cache otomatis ( dokumentasi ): Tidak

  • Perpecahan :

Membelah Contoh
'test' 116.323
'train' 697.932
  • Struktur fitur :
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=62),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (28, 28, 1) uint8
label LabelKelas int64

Visualisasi

emnist/bymerge

  • Deskripsi konfigurasi : EMNIST ByMerge

  • Ukuran dataset : 349.16 MiB

  • Di-cache otomatis ( dokumentasi ): Tidak

  • Perpecahan :

Membelah Contoh
'test' 116.323
'train' 697.932
  • Struktur fitur :
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=47),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (28, 28, 1) uint8
label LabelKelas int64

Visualisasi

emnis / seimbang

  • Deskripsi konfigurasi : EMNIST Balanced

  • Ukuran dataset : 56.63 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 18.800
'train' 112.800
  • Struktur fitur :
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=47),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (28, 28, 1) uint8
label LabelKelas int64

Visualisasi

emnis / surat

  • Deskripsi konfigurasi : EMNIST Letters

  • Ukuran dataset : 44.14 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 14.800
'train' 88.800
  • Struktur fitur :
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=37),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (28, 28, 1) uint8
label LabelKelas int64

Visualisasi

emnis/digit

  • Deskripsi konfigurasi : EMNIST Digit

  • Ukuran dataset : 120.32 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 40.000
'train' 240.000
  • Struktur fitur :
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=10),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (28, 28, 1) uint8
label LabelKelas int64

Visualisasi

emnis / mnist

  • Deskripsi konfigurasi : EMNIST MNIST

  • Ukuran dataset : 30.09 MiB

  • Di-cache otomatis ( dokumentasi ): Ya

  • Perpecahan :

Membelah Contoh
'test' 10.000
'train' 60.000
  • Struktur fitur :
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=10),
})
  • Dokumentasi fitur :
Fitur Kelas Membentuk Dtype Keterangan
fiturDict
gambar Gambar (28, 28, 1) uint8
label LabelKelas int64

Visualisasi