کیلومترنست

Kuzushiji-MNIST جایگزینی برای مجموعه داده MNIST (28x28 مقیاس خاکستری، 70000 تصویر) است که در قالب اصلی MNIST و همچنین فرمت NumPy ارائه شده است. از آنجایی که MNIST ما را به 10 کلاس محدود می کند، هنگام ایجاد Kuzushiji-MNIST یک کاراکتر را برای نمایش هر یک از 10 ردیف هیراگانا انتخاب کردیم.

شکاف مثال ها
'test' 10000
'train' 60000
  • ساختار ویژگی :
FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=uint8),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=10),
})
  • مستندات ویژگی :
ویژگی کلاس شکل نوع D شرح
FeaturesDict
تصویر تصویر (28، 28، 1) uint8
برچسب ClassLabel int64

تجسم

  • نقل قول :
@online{clanuwat2018deep,
  author       = {Tarin Clanuwat and Mikel Bober-Irizar and Asanobu Kitamoto and Alex Lamb and Kazuaki Yamamoto and David Ha},
  title        = {Deep Learning for Classical Japanese Literature},
  date         = {2018-12-03},
  year         = {2018},
  eprintclass  = {cs.CV},
  eprinttype   = {arXiv},
  eprint       = {cs.CV/1812.01718},
}