update

dataset_ini.py

@@ -0,0 +1,49 @@
+import pandas as pd
+from datasets import Dataset, DatasetDict, concatenate_datasets
+import os
+
+def load_data(csv_file, images_folder):
+    df = pd.read_csv(csv_file)
+    # Update the image_path column to include the full path
+    df['image_path'] = df['image_path'].apply(lambda x: os.path.join(images_folder, x))
+    dataset = Dataset.from_pandas(df)
+    return dataset
+
+def load_dataset():
+    # Define directories
+    test_real = 'datasets/test_set/real/'
+    test_fake = 'datasets/test_set/fake/'
+    
+    training_real = 'datasets/training_set/real/'
+    training_fake = 'datasets/training_set/fake/'
+    
+    # Define CSV files
+    train_csv = 'train_labels.csv'
+    test_csv = 'test_labels.csv'
+    
+    # Load datasets
+    train_dataset_real = load_data(train_csv, training_real)
+    
+    train_dataset_fake = load_data(train_csv, training_fake)
+    
+    # Combine datasets for training
+    train_dataset = concatenate_datasets([train_dataset_real, train_dataset_fake])
+    
+    test_dataset_real = load_data(test_csv, test_real)
+    test_dataset_fake = load_data(test_csv, test_fake)
+    
+    # Combine datasets for testing
+    test_dataset = concatenate_datasets([test_dataset_real, test_dataset_fake])
+    
+    # Create DatasetDict
+    dataset_dict = DatasetDict({
+        'train': train_dataset,
+        'test': test_dataset
+    })
+    
+    return dataset_dict
+
+if __name__ == "__main__":
+    dataset = load_dataset()
+    print(dataset)
+    

test_labels.csv

@@ -1,3 +1,4 @@
+image_path,label
 datasets/test_set/real/Gustave_Courbet_11.jpg,0
 datasets/test_set/real/Gustave_Courbet_12.jpg,0
 datasets/test_set/real/Gustave_Courbet_13.jpg,0

train_labels.csv

@@ -1,3 +1,4 @@
+image_path,label
 datasets/training_set/real/Alfred_Sisley_1.jpg,0
 datasets/training_set/real/Alfred_Sisley_10.jpg,0
 datasets/training_set/real/Alfred_Sisley_100.jpg,0