|
O’zbekiston respublikasi axborot texnologiyalari va kommunikasiyalarini rivojlantirish vazirligi
|
| bet | 1/3 | | Sana | 01.04.2024 | | Hajmi | 1.37 Mb. | | #184004 |
Bog'liq Amaliy ish 493, [@ABTbooks] 5000 words in a month
O’ZBEKISTON RESPUBLIKASI AXBOROT TEXNOLOGIYALARI VA KOMMUNIKASIYALARINI RIVOJLANTIRISH VAZIRLIGI
MUHAMMAD AL-XORAZMIY NOMIDAGI
TOSHKENT AXBOROT TEXNOLOGIYALARI UNIVERSITETI
DASTURIY INJINIRING FAKULTETI
Amaliy ish
Bajardi: Tekshirdi:
Тoshkent – 2024
# importing the general dependencies
import pandas as pdimport numpy as npimport matplotlib.pyplot as pltimport cv2import osimport warnings
warnings.filterwarnings('ignore')
In [33]:
# Lets check what we have in our dataset
dataset_root = '/kaggle/input/affectnet-yolo-format/YOLO_format/'
for root, directories, files in os.walk(dataset_root):
# Do something with the current directory path 'root'
print("Current directory:", root)
# Do something with the list of subdirectories 'directories'
print("Subdirectories:", directories)
# Do something with the list of files 'files'
print("Files:", len(files))
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/
Subdirectories: ['valid', 'test', 'train']
Files: 1
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/valid
Subdirectories: ['labels', 'images']
Files: 0
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/valid/labels
Subdirectories: []
Files: 5406
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/valid/images
Subdirectories: []
Files: 5406
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/test
Subdirectories: ['labels', 'images']
Files: 0
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/test/labels
Subdirectories: []
Files: 2755
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/test/images
Subdirectories: []
Files: 2755
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/train
Subdirectories: ['labels', 'images']
Files: 0
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/train/labels
Subdirectories: []
Files: 17101
Current directory: /kaggle/input/affectnet-yolo-format/YOLO_format/train/images
Subdirectories: []
Files: 17101
In [34]:
# Lets observe some images from three folders
# Define the paths to the image foldersimage_folders = [
"/kaggle/input/affectnet-yolo-format/YOLO_format/valid/images",
"/kaggle/input/affectnet-yolo-format/YOLO_format/test/images",
"/kaggle/input/affectnet-yolo-format/YOLO_format/train/images"]
# Define the number of images you want to display from each foldernum_images_per_folder = 5
# Create a subplot gridfig, axes = plt.subplots(len(image_folders), num_images_per_folder, figsize=(15, 10))
# Loop through each image folderfor i, folder in enumerate(image_folders):
image_files = os.listdir(folder)[:num_images_per_folder] # Get the list of image files
for j, image_file in enumerate(image_files):
image_path = os.path.join(folder, image_file)
try:
# Open and display the image
img = plt.imread(image_path)
axes[i, j].imshow(img)
axes[i, j].axis('off')
except Exception as e:
print(f"Error opening image: {image_path}")
# Adjust layoutplt.tight_layout()plt.show()
In [35]:
# Lets observe an individual image with its corresponding labellabels_map = {
'0' : 'Anger',
'1' : 'Contempt',
'2' : 'Disgust',
'3' : 'Fear',
'4' : 'Happy',
'5' : 'Neutral',
'6' : 'Sad',
'7' : 'Surprise'}
random_number = 333train_images_path = '/kaggle/input/affectnet-yolo-format/YOLO_format/train/images'random_image_path = os.path.join(train_images_path, os.listdir(train_images_path)[random_number])random_image = plt.imread(random_image_path)
train_labels_path = '/kaggle/input/affectnet-yolo-format/YOLO_format/train/labels'random_label_file = random_image_path.split('/')[-1].split('.')[0] + '.txt'random_label_path = os.path.join(train_labels_path, random_label_file)
with open(random_label_path, 'r') as file:
# first character is the label
first_character = file.read(1)image_class = labels_map[first_character]
print(random_label_path)print(random_image_path)
plt.imshow(random_image)plt.title(f'{image_class}')plt.show()
/kaggle/input/affectnet-yolo-format/YOLO_format/train/labels/ffhq_5410.txt
/kaggle/input/affectnet-yolo-format/YOLO_format/train/images/ffhq_5410.png
In [36]:
# Class frequenciesclass_frequencies = np.array([])
# Define the paths to the image folderslabel_folders = [
"/kaggle/input/affectnet-yolo-format/YOLO_format/valid/labels",
"/kaggle/input/affectnet-yolo-format/YOLO_format/test/labels",
"/kaggle/input/affectnet-yolo-format/YOLO_format/train/labels"]
# Iterate over each label directory (valid, test, train)for label_dir in label_folders:
# Iterate over each .txt file in the labels directory
for file_name in os.listdir(label_dir):
if file_name.endswith('.txt'):
file_path = os.path.join(label_dir, file_name)
# Read the first character of the file
with open(file_path, 'r') as file:
first_character = file.read(1)
class_frequencies = np.append(class_frequencies, first_character)
In [37]:
unique, counts = np.unique(class_frequencies, return_counts=True)unique_mapped = [labels_map[x] for x in unique]
# Create the bar plotplt.bar(unique_mapped, counts)
# Add labels and titleplt.xlabel('Number')plt.ylabel('Frequency')plt.title('Frequency of Numbers')plt.xticks(rotation=45)
# Show the plotplt.show()
|
| |
