fix: index from 0

ResNet-CIFAR10/Classification-backdoor/dataset/index.json

@@ -1,5 +1,6 @@
 {
     "train": [
+        0,
         1,
         2,
         3,
@@ -49998,10 +49999,10 @@
         49996,
         49997,
         49998,
-        49999,
-        50000
+        49999
     ],
     "test": [
+        50000,
         50001,
         50002,
         50003,
@@ -60000,8 +60001,7 @@
         59996,
         59997,
         59998,
-        59999,
-        60000
+        59999
     ],
     "validation": []
 }

ResNet-CIFAR10/Classification-backdoor/scripts/create_index.py

@@ -3,8 +3,8 @@ import os
 
 # 创建完整的索引
 index_dict = {
-    "train": list(range(1, 50001)),  # 从1到50000的训练索引
-    "test": list(range(50001, 60001)),  # 从50001到60000的测试索引
+    "train": list(range( 50000)),  # 从1到50000的训练索引
+    "test": list(range(50000, 60000)),  # 从50001到60000的测试索引
     "validation": []  # 空验证集
 }
 

ResNet-CIFAR10/Classification-backdoor/scripts/get_raw_data.py

@@ -88,10 +88,10 @@ def save_images_from_cifar10_with_backdoor(dataset_path, save_dir):
             img_backdoor[-trigger_size:, -trigger_size:, :] = 255
             # 保存带触发器的图像
             img_backdoor_pil = Image.fromarray(img_backdoor)
-            img_backdoor_pil.save(os.path.join(save_dir, f"{i+1}.png"))
+            img_backdoor_pil.save(os.path.join(save_dir, f"{i}.png"))
 
         else:
-            img_pil.save(os.path.join(save_dir, f"{i+1}.png"))
+            img_pil.save(os.path.join(save_dir, f"{i}.png"))
     
     print(f"完成! {len(all_data)} 张原始图像已保存到 {save_dir}")
 

ResNet-CIFAR10/Classification-backdoor/scripts/get_representation.py

@@ -235,8 +235,8 @@ class time_travel_saver:
             
             # 创建索引字典
             index_dict = {
-                "train": list(range(1,50001)),  # 所有数据默认为训练集
-                "test": list(range(50001, 60001)),  # 测试集索引从50000到59999
+                "train": list(range(50000)),  # 所有数据默认为训练集
+                "test": list(range(50000, 60000)),  # 测试集索引从50000到59999
                 "validation": []   # 初始为空
             }
             

ResNet-CIFAR10/Classification-normal/dataset/index.json

@@ -1,5 +1,6 @@
 {
     "train": [
+        0,
         1,
         2,
         3,
@@ -49998,8 +49999,7 @@
         49996,
         49997,
         49998,
-        49999,
-        50000
+        49999
     ],
     "test": [],
     "validation": []

ResNet-CIFAR10/Classification-normal/scripts/get_raw_data.py

@@ -62,7 +62,7 @@ def save_images_from_cifar10(dataset_path, save_dir):
     print(f"保存 {len(all_data)} 张图像...")
     for i, (img, label) in enumerate(tqdm(zip(all_data, all_labels), total=len(all_data))):
         img = Image.fromarray(img)
-        img.save(os.path.join(save_dir, f"{i+1}.png"))
+        img.save(os.path.join(save_dir, f"{i}.png"))
     
     print(f"完成! {len(all_data)} 张图像已保存到 {save_dir}")
 

ResNet-CIFAR10/Classification-normal/scripts/get_representation.py

@@ -231,7 +231,7 @@ class time_travel_saver:
             
             # 创建数据集索引
             num_samples = len(labels)
-            indices = list(range(1,num_samples+1))
+            indices = list(range(num_samples))
             
             # 创建索引字典
             index_dict = {