letter_recognition / README.md
pittawat's picture
Update README.md
19a2603
metadata
dataset_info:
  features:
    - name: image
      dtype: image
    - name: label
      dtype:
        class_label:
          names:
            '0': A
            '1': B
            '2': C
            '3': D
            '4': E
            '5': F
            '6': G
            '7': H
            '8': I
            '9': J
            '10': K
            '11': L
            '12': M
            '13': 'N'
            '14': O
            '15': P
            '16': Q
            '17': R
            '18': S
            '19': T
            '20': U
            '21': V
            '22': W
            '23': X
            '24': 'Y'
            '25': Z
  splits:
    - name: train
      num_bytes: 22453522
      num_examples: 26000
    - name: test
      num_bytes: 2244964.8
      num_examples: 2600
  download_size: 8149945
  dataset_size: 24698486.8
task_categories:
  - image-classification
language:
  - en
size_categories:
  - 1K<n<10K

Dataset Card for "letter_recognition"

Images in this dataset was generated using the script defined below. The original dataset in CSV format and more information of the original dataset is available at A-Z Handwritten Alphabets in .csv format.

import os
import pandas as pd
import matplotlib.pyplot as plt

CHARACTER_COUNT = 26

data = pd.read_csv('./A_Z Handwritten Data.csv')
mapping = {str(i): chr(i+65) for i in range(26)}

def generate_dataset(folder, end, start=0):
    if not os.path.exists(folder):
        os.makedirs(folder)
        print(f"The folder '{folder}' has been created successfully!")
    else:
        print(f"The folder '{folder}' already exists.")

    for i in range(CHARACTER_COUNT):
        dd = data[data['0']==i]
        for j in range(start, end):
            ddd = dd.iloc[j]
            x = ddd[1:].values
            x = x.reshape((28, 28))
            plt.axis('off')
            plt.imsave(f'{folder}/{mapping[str(i)]}_{j}.jpg', x, cmap='binary')
            
generate_dataset('./train', 1000)
generate_dataset('./test', 1100, 1000)