行人重识别深度学习踩坑笔记

1. enumerate(dataloader）报错

train函数中在for循环中进行enumerate（dataloader）时报错

Traceback (most recent call last):
  File "G:/行人重识别学习/try/train_class.py", line 245, in <module>
    main()
  File "G:/行人重识别学习/try/train_class.py", line 228, in main
    train(epoch, model, criterion_class, optimizer, trainloader, use_gpu)
  File "G:/行人重识别学习/try/train_class.py", line 232, in train
    for batch_idx, (imgs, pids, _) in enumerate(trainloader):
  File "G:\anaconda1\lib\site-packages\torch\utils\data\dataloader.py", line 345, in __next__
    data = self._next_data()
  File "G:\anaconda1\lib\site-packages\torch\utils\data\dataloader.py", line 384, in _next_data
    index = self._next_index()  # may raise StopIteration
  File "G:\anaconda1\lib\site-packages\torch\utils\data\dataloader.py", line 339, in _next_index
    return next(self._sampler_iter)  # may raise StopIteration
  File "G:\anaconda1\lib\site-packages\torch\utils\data\sampler.py", line 200, in __iter__
    for idx in self.sampler:
  File "G:\anaconda1\lib\site-packages\torch\utils\data\sampler.py", line 62, in __iter__
    return iter(range(len(self.data_source)))
TypeError: 'list' object cannot be interpreted as an integer

错误原因：由于行人重识别不同于一般的深度学习，需要自己构建数据集结构，数据集类有三个关键函数：init、len和getitem。
在重写dataset过程中，出现一个低级错误，__len__函数的返回值没加len（）

# 数据库类三个必备函数  init len getitem
class ImageDataset(Dataset):
    def __init__(self, dataset, transform=None):
        self.dataset = dataset
        self.transform = transform

    def __len__(self):
        return self.dataset   #这行出大问题  len（）忘记加了
		return len(self.dataset)  #正确写法
    def __getitem__(self, index):
        img_path, pid, camid = self.dataset[index]
        img = read_image(img_path)
        if self.transform is not None:
            img = self.transform(img)
        return img, pid, camid

2. torch.Resize((height,width))出错

少了个括号，占用了别的参数的位置

transform_test = T.Compose([
        # T.Resize(args.height, args.width),这行出错
        T.Resize((args.height, args.width))   #
        T.ToTensor(),
        T.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
    ])

Traceback (most recent call last):
  File "G:/行人重识别学习/try/train_class.py", line 305, in <module>
    main()
  File "G:/行人重识别学习/try/train_class.py", line 227, in main
    test(model, queryloader, galleryloader, use_gpu)
  File "G:/行人重识别学习/try/train_class.py", line 88, in test
    for batch_idx, (imgs, pids, camids) in enumerate(queryloader):
  File "G:\anaconda1\lib\site-packages\torch\utils\data\dataloader.py", line 345, in __next__
    data = self._next_data()
  File "G:\anaconda1\lib\site-packages\torch\utils\data\dataloader.py", line 385, in _next_data
    data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
  File "G:\anaconda1\lib\site-packages\torch\utils\data\_utils\fetch.py", line 44, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "G:\anaconda1\lib\site-packages\torch\utils\data\_utils\fetch.py", line 44, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "G:\行人重识别学习\try\data_loader.py", line 40, in __getitem__
    img = self.transform(img)
  File "G:\anaconda1\lib\site-packages\torchvision\transforms\transforms.py", line 70, in __call__
    img = t(img)
  File "G:\anaconda1\lib\site-packages\torchvision\transforms\transforms.py", line 207, in __call__
    return F.resize(img, self.size, self.interpolation)
  File "G:\anaconda1\lib\site-packages\torchvision\transforms\functional.py", line 250, in resize
    return img.resize((ow, oh), interpolation)
  File "G:\anaconda1\lib\site-packages\PIL\Image.py", line 1742, in resize
    raise ValueError("unknown resampling filter")
ValueError: unknown resampling filter

3. IndexError: tensors used as indices must be long, byte or bool tensors

p_local_features = local_features[p_inds]   # N*H * C  
# p_inds是一个floatTensor 利用.long()转化为整形
# IndexError: tensors used as indices must be long, byte or bool tensors
# 要将tensor作为索引使用，tensor必须是一个long byte或者bool tensor
p_inds, n_inds = p_inds.long(), n_inds.long()

4.损失函数输入形状[64,1]和标签[64]不匹配

d_loss_fake = criterion(fake_out,fake_label)           #修改前
d_loss_fake = criterion(fake_out.squeeze(-1),fake_label)  # 修改后

计算loss时出现的错误，预测的batch维度与真实的batch维度不同，要压缩一个，具体压缩那个看输出信息
D:\ProgramData\Anaconda3\envs\pytorch\lib\site-packages\torch\nn\modules\loss.py:443: UserWarning: Using a target size (torch.Size([10])) that is different to the input size (torch.Size([1, 10])). This will likely lead to incorrect results due to broadcasting. Please ensure they have the same size.
return F.mse_loss(input, target, reduction=self.reduction)