Serbian Sign Language Alphabet (SSL) - NumPy DATASET

All 30 Serbian alphabets | Distributed in Folders and Described with key points

This SSL Dataset can be used to create Machine Learning models: Games, Educational Material, Communication Software, ...

SSL Dataset Contents (file count):

A = 211 |Б = 264 |В = 265 |Г = 220 |Д = 215 |Ђ = 222 |Е = 231 |Ж = 225 |З = 224 |И = 240 |
Ј = 247 |К = 205 |Л = 451 |Љ = 191 |М = 226 |Н = 227 |Њ = 208 |О = 409 |П = 363 |Р = 294 |
С = 296 |Т = 302 |Ћ = 269 |У = 200 |Ф = 184 |Х = 303 |Ц = 285 |Ч = 262 |Џ = 256 |Ш = 257 |

Figure 1. NumPy file structure used in SSL dataset

How to load a NumPy file from SSL Dataset and use it?

Dataset can be used for creating machine learning models.

import numpy as np
import os
import time

def convert_npy_file_data_to_csv(file_name):
    numpy_array = np.load(file_name)
    total_values = len(numpy_array)
    ELEMENTS_IN_ONE_FRAME = 33*4 + 21*3 + 21*3
    counter = 0
    frame_number = 0
    all_frames = []
    for j in range(int(len(numpy_array)/(ELEMENTS_IN_ONE_FRAME))):
        frame_number += 1
        data_row = []
        for i in range(ELEMENTS_IN_ONE_FRAME):
            data_row.append(numpy_array[i+j*ELEMENTS_IN_ONE_FRAME])
        all_frames.append(data_row)
    print(all_frames)
    np.savetxt(f"{file_name}.csv", all_frames, delimiter=",")  # EXPORT DATA TO A .csv FILE
    print(f"Successfully saved {file_name}.csv file.")
    return

if __name__ == "__main__":
    file_name = "..\\SSL Program\\D\\17039342851960452-XVID.npy"
    convert_npy_file_data_to_csv(file_name)