增强后的图片的split和stack

[creater-zq]

您好:

• 如果我的batch_size = 4，请问是不是需要先split拆分为4个tensor,然后每个tensor当独做normlization，之后在stack为一个tensor，再送个分割网络？谢谢
  

• 是否可以采用下面的方式，这样batch_size可以任意？

normal = transforms.Compose([transforms.Normalize(mean=[0.485, 0.456, 0.406],
                                                  std=[0.229, 0.224, 0.225])])

enhanced_image = torch.Tensor(4, 3, 512, 512)
new_image = None
for idx, cur_img in enumerate(enhanced_image):
    if idx == 0:
        new_image = normal(cur_img).unsqueeze(dim=0)
    else:
        cur_img = normal(cur_img).unsqueeze(dim=0)
        new_image = torch.concat((new_image, cur_img), dim=0)