昇思25天学习打卡营第15天|ResNet50迁移学习

迁移学习

通常的做法是在非常大的数据集（如ImageNet，其中包含120万张图片和1000个类别）上对卷积神经网络预训练，再在实际任务中使用这个网络或再进行微调。

有两种主要的迁移学习方案：
1. 预训练模型作为固定特征提取器：自己训练或者拿别人的预训练模型，移除最后的全连接层（这层是用来对图片分类的），把这个网络作为固定特征提取器。再为新数据集替换和重新训练分类器。
2. 微调ConvNet：在1的基础上，还继续通过反向传播训练网络权重。这里可以保持某些底层的参数固定，只微调一些更高级别的部分。

微调分成多个场景：
1. 新数据集量级小，和原数据集相似：由于数据集较小，因此再去微调就会有过拟合的风险。这里最好是直接训练一个线性分类器
2. 新数据集量级大，和原数据集相似：数据集较大时没有过拟合风险，因此可以对整个网络微调训练。
3. 新数据量级小，不同于原数据集：小数据集最好只训练线性分类器，而数据集差异很大，最好就从网络顶部训练分类器，因为网络中包含了很多训练数据集的特征。因此这里选择从网络的早期某处激活中训练SVM分类器。
4. 新数据量级大，不同于原数据集：可以对整个网络微调

数据加载

使用分类狼与狗的Canidae数据集，该数据集图像来自ImageNet，每个分类包含120张训练图和30张验证图。

batch_size = 18                             # 批量大小
image_size = 224                            # 训练图像空间大小
num_epochs = 5                             # 训练周期数
lr = 0.001                                  # 学习率
momentum = 0.9                              # 动量
workers = 4                                 # 并行线程个数

创建训练数据集

def create_dataset_canidae(dataset_path, usage):
    data_set = ds.ImageFolderDataset(dataset_path,
                                     num_parallel_workers=workers,
                                     shuffle=True,)

    # 数据增强操作
    mean = [0.485 * 255, 0.456 * 255, 0.406 * 255]
    std = [0.229 * 255, 0.224 * 255, 0.225 * 255]
    scale = 32

    if usage == "train":
        #训练集则随机剪裁、解码、缩放、水平翻转、正则化、改变通道顺序。
        trans = [
            vision.RandomCropDecodeResize(size=image_size, scale=(0.08, 1.0), ratio=(0.75, 1.333)),
            vision.RandomHorizontalFlip(prob=0.5),
            vision.Normalize(mean=mean, std=std),
            vision.HWC2CHW()
        ]
    else:
        # 解码、重定义大小、中心裁剪、正则化、改变通道顺序
        trans = [
            vision.Decode(),
            vision.Resize(image_size + scale),
            vision.CenterCrop(image_size),
            vision.Normalize(mean=mean, std=std),
            vision.HWC2CHW()
        ]


    # 数据映射操作
    data_set = data_set.map(
        operations=trans,
        input_columns='image',
        num_parallel_workers=workers)


    # 批量操作
    data_set = data_set.batch(batch_size)

    return data_set


dataset_train = create_dataset_canidae(data_path_train, "train")
step_size_train = dataset_train.get_dataset_size()

dataset_val = create_dataset_canidae(data_path_val, "val")
step_size_val = dataset_val.get_dataset_size()

模型搭建

构建一个基础的残差块

class ResidualBlockBase(nn.Cell):
    expansion: int = 1  # 最后一个卷积核数量与第一个卷积核数量相等

    def __init__(self, in_channel: int, out_channel: int,
                 stride: int = 1, norm: Optional[nn.Cell] = None,
                 down_sample: Optional[nn.Cell] = None) -> None:
        super(ResidualBlockBase, self).__init__()
        if not norm:
            self.norm = nn.BatchNorm2d(out_channel)
        else:
            self.norm = norm
		# 创建两个3*3的卷积层
        self.conv1 = nn.Conv2d(in_channel, out_channel,
                               kernel_size=3, stride=stride,
                               weight_init=weight_init)
        self.conv2 = nn.Conv2d(in_channel, out_channel,
                               kernel_size=3, weight_init=weight_init)
        self.relu = nn.ReLU()
        self.down_sample = down_sample

    def construct(self, x):
        identity = x  # shortcuts分支（跳跃连接）

        out = self.conv1(x)  # 主分支第一层：3*3卷积层
        out = self.norm(out)
        out = self.relu(out)
        out = self.conv2(out)  # 主分支第二层：3*3卷积层
        out = self.norm(out)

        if self.down_sample is not None:
            identity = self.down_sample(x)
        out += identity  # 输出为主分支与shortcuts之和
		# 加在一起后再经过ReLU激活函数
        out = self.relu(out)

        return out

构建残差层

class ResidualBlock(nn.Cell):
    expansion = 4  # 最后一个卷积核的数量是第一个卷积核数量的4倍

    def __init__(self, in_channel: int, out_channel: int,
                 stride: int = 1, down_sample: Optional[nn.Cell] = None) -> None:
        super(ResidualBlock, self).__init__()

        self.conv1 = nn.Conv2d(in_channel, out_channel,
                               kernel_size=1, weight_init=weight_init)
        self.norm1 = nn.BatchNorm2d(out_channel)
        self.conv2 = nn.Conv2d(out_channel, out_channel,
                               kernel_size=3, stride=stride,
                               weight_init=weight_init)
        self.norm2 = nn.BatchNorm2d(out_channel)
        self.conv3 = nn.Conv2d(out_channel, out_channel * self.expansion,
                               kernel_size=1, weight_init=weight_init)
        self.norm3 = nn.BatchNorm2d(out_channel * self.expansion)

        self.relu = nn.ReLU()
        self.down_sample = down_sample

    def construct(self, x):

        identity = x  # shortscuts分支

        out = self.conv1(x)  # 主分支第一层：1*1卷积层
        out = self.norm1(out)
        out = self.relu(out)
        out = self.conv2(out)  # 主分支第二层：3*3卷积层
        out = self.norm2(out)
        out = self.relu(out)
        out = self.conv3(out)  # 主分支第三层：1*1卷积层
        out = self.norm3(out)

        if self.down_sample is not None:
            identity = self.down_sample(x)

        out += identity  # 输出为主分支与shortcuts之和
        out = self.relu(out)

        return out

构建堆叠的残差块。

def make_layer(last_out_channel, block: Type[Union[ResidualBlockBase, ResidualBlock]],
               channel: int, block_nums: int, stride: int = 1):
    down_sample = None  # shortcuts分支

    if stride != 1 or last_out_channel != channel * block.expansion:
	# 构建 下采样模块：1个卷积层+批正则化层
	# 下采样用于减少特征图的空间分辨率，同时增加特征的抽象程度
        down_sample = nn.SequentialCell([
            nn.Conv2d(last_out_channel, channel * block.expansion,
                      kernel_size=1, stride=stride, weight_init=weight_init),
            nn.BatchNorm2d(channel * block.expansion, gamma_init=gamma_init)
        ])

    layers = []
    layers.append(block(last_out_channel, channel, stride=stride, down_sample=down_sample))

    in_channel = channel * block.expansion
    
    # 堆叠残差网络
    for _ in range(1, block_nums):
        layers.append(block(in_channel, channel))

    return nn.SequentialCell(layers)

构建ResNet网络

class ResNet(nn.Cell):
    def __init__(self, block: Type[Union[ResidualBlockBase, ResidualBlock]],
                 layer_nums: List[int], num_classes: int, input_channel: int) -> None:
        super(ResNet, self).__init__()

        self.relu = nn.ReLU()
        # 第一个卷积层，输入channel为3（彩色图像），输出channel为64
        self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, weight_init=weight_init)
	# 使用批正则化
        self.norm = nn.BatchNorm2d(64)
        # 最大池化层，缩小图片的尺寸
        self.max_pool = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode='same')
        # 各个残差网络结构块定义，
        self.layer1 = make_layer(64, block, 64, layer_nums[0])
        self.layer2 = make_layer(64 * block.expansion, block, 128, layer_nums[1], stride=2)
        self.layer3 = make_layer(128 * block.expansion, block, 256, layer_nums[2], stride=2)
        self.layer4 = make_layer(256 * block.expansion, block, 512, layer_nums[3], stride=2)
        # 平均池化层
        self.avg_pool = nn.AvgPool2d()
        # flattern层
        self.flatten = nn.Flatten()
        # 全连接层
        self.fc = nn.Dense(in_channels=input_channel, out_channels=num_classes)

    def construct(self, x):

        x = self.conv1(x)
        x = self.norm(x)
        x = self.relu(x)
        x = self.max_pool(x)
		# 四个残差块
        x = self.layer1(x)
        x = self.layer2(x)
        x = self.layer3(x)
        x = self.layer4(x)
		# 收尾：平均池化层、展平、全连接层
        x = self.avg_pool(x)
        x = self.flatten(x)
        x = self.fc(x)

        return x

真正开始构建，可以选择使用预训练模型。

def _resnet(model_url: str, block: Type[Union[ResidualBlockBase, ResidualBlock]],
            layers: List[int], num_classes: int, pretrained: bool, pretrianed_ckpt: str,
            input_channel: int):
    model = ResNet(block, layers, num_classes, input_channel)

    if pretrained:
        # 加载预训练模型
        download(url=model_url, path=pretrianed_ckpt, replace=True)
        param_dict = load_checkpoint(pretrianed_ckpt)
        load_param_into_net(model, param_dict)

    return model


def resnet50(num_classes: int = 1000, pretrained: bool = False):
      resnet50_url = "https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/models/application/resnet50_224_new.ckpt"
    resnet50_ckpt = "./LoadPretrainedModel/resnet50_224_new.ckpt"
    return _resnet(resnet50_url, ResidualBlock, [3, 4, 6, 3], num_classes,
                   pretrained, resnet50_ckpt, 2048)

固定特征训练

这里冻结最后一层意外的所有网络层，使用requires_grad == False 冻结参数

import mindspore as ms
import matplotlib.pyplot as plt
import os
import time

# 实例化网络
net_work = resnet50(pretrained=True)

# 全连接层输入层的大小
in_channels = net_work.fc.in_channels
# 输出通道数大小为狼狗分类数2
head = nn.Dense(in_channels, 2)
# 重置全连接层
net_work.fc = head

# 平均池化层kernel size为7
avg_pool = nn.AvgPool2d(kernel_size=7)
# 重置平均池化层
net_work.avg_pool = avg_pool

# 冻结除最后一层外的所有参数
for param in net_work.get_parameters():
    if param.name not in ["fc.weight", "fc.bias"]:
        param.requires_grad = False

# 定义优化器和损失函数
opt = nn.Momentum(params=net_work.trainable_params(), learning_rate=lr, momentum=0.5)
loss_fn = nn.SoftmaxCrossEntropyWithLogits(sparse=True, reduction='mean')


def forward_fn(inputs, targets):
    logits = net_work(inputs)
    loss = loss_fn(logits, targets)

    return loss

grad_fn = ms.value_and_grad(forward_fn, None, opt.parameters)

def train_step(inputs, targets):
    loss, grads = grad_fn(inputs, targets)
    opt(grads)
    return loss

# 实例化模型
model1 = train.Model(net_work, loss_fn, opt, metrics={"Accuracy": train.Accuracy()})

训练和评估

import mindspore as ms
import matplotlib.pyplot as plt
import os
import time
# 开始循环训练
print("Start Training Loop ...")

best_acc = 0

# 按照定好的epoch数进行迭代
for epoch in range(num_epochs):
    losses = []
    net_work.set_train()

    # 为每轮训练读入数据
    for i, (images, labels) in enumerate(data_loader_train):
        labels = labels.astype(ms.int32)
        loss = train_step(images, labels)
        losses.append(loss)

    # 每个epoch结束后，验证准确率
    acc = model1.eval(dataset_val)['Accuracy']
    
    print("-" * 20)
    print("Epoch: [%3d/%3d], Average Train Loss: [%5.3f], Accuracy: [%5.3f]" % (
        epoch+1, num_epochs, sum(losses)/len(losses), acc
    ))
   
   # 保存下来准确度最佳的模型
    if acc > best_acc:
        best_acc = acc
        if not os.path.exists(best_ckpt_dir):
            os.mkdir(best_ckpt_dir)
        ms.save_checkpoint(net_work, best_ckpt_path)

print("=" * 80)
print(f"End of validation the best Accuracy is: {best_acc: 5.3f}, "
      f"save the best ckpt file in {best_ckpt_path}", flush=True)

预测

import matplotlib.pyplot as plt
import mindspore as ms

# 可视化预测结果
def visualize_model(best_ckpt_path, val_ds):
    net = resnet50()
    # 全连接层输入层的大小
    in_channels = net.fc.in_channels
    # 输出通道数大小为狼狗分类数2
    head = nn.Dense(in_channels, 2)
    # 重置全连接层
    net.fc = head
    # 平均池化层kernel size为7
    avg_pool = nn.AvgPool2d(kernel_size=7)
    # 重置平均池化层
    net.avg_pool = avg_pool
    # 加载模型参数
    param_dict = ms.load_checkpoint(best_ckpt_path)
    ms.load_param_into_net(net, param_dict)
    model = train.Model(net)
    # 加载验证集的数据进行验证
    data = next(val_ds.create_dict_iterator())
    images = data["image"].asnumpy()
    labels = data["label"].asnumpy()
    class_name = {0: "dogs", 1: "wolves"}
    # 预测图像类别
    output = model.predict(ms.Tensor(data['image']))
    pred = np.argmax(output.asnumpy(), axis=1)

    # 显示图像及图像的预测值
    plt.figure(figsize=(5, 5))
    for i in range(4):
        plt.subplot(2, 2, i + 1)
        # 若预测正确，显示为蓝色；若预测错误，显示为红色
        color = 'blue' if pred[i] == labels[i] else 'red'
        plt.title('predict:{}'.format(class_name[pred[i]]), color=color)
        picture_show = np.transpose(images[i], (1, 2, 0))
        mean = np.array([0.485, 0.456, 0.406])
        std = np.array([0.229, 0.224, 0.225])
        picture_show = std * picture_show + mean
        picture_show = np.clip(picture_show, 0, 1)
        plt.imshow(picture_show)
        plt.axis('off')

    plt.show()

最后得到模型的预测结果。

总结

本章依然使用了ResNet50网络模型，完成了狼狗图片分类任务。

打卡凭证