Create data_loader_new.py

utils/data_loader_new.py

@@ -0,0 +1,76 @@
+'''
+for debug
+'''
+
+import torch
+from torch.utils.data import Dataset, DataLoader
+from torchvision import transforms
+from PIL import Image
+import csv
+
+preprocess = transforms.Compose([
+    transforms.ToTensor(),
+    transforms.Normalize((0.485, 0.456, 0.406),(0.229, 0.224, 0.225))
+])
+
+def default_loader(img_path,csv_path):
+    img_tensor_list = []
+    target_tensor_list = []
+    csv_reader = csv.reader(open(csv_path,'r'))
+    for num, row in enumerate(csv_reader):
+        for i in range(len(row) - 1):
+            if i == 0:
+                img_pil = Image.open(img_path + row[i])
+                img_pil = img_pil.resize((224, 224))
+                img_tensor = preprocess(img_pil)
+                img_tensor = img_tensor.expand(1, 3, 224, 224)
+                bag_tensor = img_tensor
+                label = []
+                label.append(int(row[-1]))
+            else:
+                img_pil = Image.open(img_path + row[i])
+                img_pil = img_pil.resize((224, 224))
+                img_tensor = preprocess(img_pil)
+                img_tensor = img_tensor.expand(1, 3, 224, 224)
+                bag_tensor = torch.cat((bag_tensor,img_tensor),0)                
+
+        bag_tensor_label = torch.LongTensor(label)
+        img_tensor_list.append(bag_tensor)
+        target_tensor_list.append(bag_tensor_label)
+    print(img_tensor_list[0].size()) # [15,3,224,224]
+    print(target_tensor_list[0].size()) # [1]
+    return img_tensor_list, target_tensor_list
+
+class ThyroidDataset(Dataset):
+    def __init__(self, loader=default_loader, img_path=None, csv_path=None):
+        self.img_path = img_path
+        self.csv_path = csv_path
+        csv_reader = csv.reader(open(csv_path,'r'))
+        self.instance_list = list(csv_reader)
+        self.length = len(self.instance_list)
+
+    def __getitem__(self, index):
+
+        row = self.instance_list[index]
+        for i in range(len(row) - 1):
+            if i == 0:
+                img_pil = Image.open(self.img_path + row[i])
+                img_pil = img_pil.resize((224, 224))
+                img_tensor = preprocess(img_pil)
+                img_tensor = img_tensor.expand(1, 3, 224, 224)
+                bag_tensor = img_tensor
+                label = []
+                label.append(int(row[-1]))
+            else:
+                img_pil = Image.open(self.img_path + row[i])
+                img_pil = img_pil.resize((224, 224))
+                img_tensor = preprocess(img_pil)
+                img_tensor = img_tensor.expand(1, 3, 224, 224)
+                bag_tensor = torch.cat((bag_tensor,img_tensor),0)                
+
+        bag_tensor_label = torch.LongTensor(label)
+        return bag_tensor, bag_tensor_label
+
+    def __len__(self):
+#         return len(self.img_tensor_list)
+        return self.length