Pytorch自定义randomcrop用于语义分割

如何解决Pytorch自定义randomcrop用于语义分割

我正在尝试实现自定义数据集加载器。首先，我以（0.98，1,1）之间相同的比例调整图像和标签的大小，然后我以相同的参数随机裁剪图像和标签，以便将它们输入到NN中。但是，我从pytorch函数中得到一个错误。这是我的代码：

class RandomCrop(object):

    def __init__(self,size,padding=None,pad_if_needed=True,fill=0,padding_mode='constant'):
        
        self.size = size
        self.padding = padding
        self.pad_if_needed = pad_if_needed
        self.fill = fill
        self.padding_mode = padding_mode

    @staticmethod
    def get_params(img,output_size):
        
        w,h = img.size
        th,tw = output_size
        if w == tw and h == th:
            return 0,h,w

        i = random.randint(0,h - th)
        j = random.randint(0,w - tw)
        return i,j,th,tw

    def __call__(self,data):
     
        img,mask = data["image"],data["mask"]
       

        # pad the width if needed
        if self.pad_if_needed and img.size[0] < self.size[1]:
            img = F.pad(img,(self.size[1] - img.size[0],0),self.fill,self.padding_mode)
            mask = F.pad(mask,(self.size[1] - mask.size[0],self.padding_mode)
        # pad the height if needed
        if self.pad_if_needed and img.size[1] < self.size[0]:
            img = F.pad(img,(0,self.size[0] - img.size[1]),self.size[0] - mask.size[1]),self.padding_mode)
       
        i,w = self.get_params(img,self.size)
        crop_image = transforms.functional.crop(img,i,w)
        crop_mask = transforms.functional.crop(mask,w)

        return{"image": crop_image,"mask": crop_mask }

这是错误：

AttributeError: 'Image' object has no attribute 'dim'

解决方法

我错误地导入了nn.functional.pad而不是transforms.functional.pad。更改后，一切进展顺利