update shallow copy method to allow variable size tensor input (#1111)

Author: buoyancy99

torchvision/models/detection/transform.py

@@ -30,7 +30,7 @@ def __init__(self, min_size, max_size, image_mean, image_std):
         self.image_std = image_std
 
     def forward(self, images, targets=None):
-        images = images[:]
+        images = [img for img in images]
         for i in range(len(images)):
             image = images[i]
             target = targets[i] if targets is not None else targets