pytorch + visdom CNN处理自建图片数据集的方法

时间:2021-05-22

环境

系统:win10

cpu:i7-6700HQ

gpu:gtx965m

python : 3.6

pytorch :0.3

数据下载

来源自Sasank Chilamkurthy 的教程; 数据:下载链接。

下载后解压放到项目根目录:


数据集为用来分类 蚂蚁和蜜蜂。有大约120个训练图像,每个类有75个验证图像。

数据导入

可以使用 torchvision.datasets.ImageFolder(root,transforms) 模块 可以将 图片转换为 tensor。

先定义transform:

ata_transforms = { 'train': transforms.Compose([ # 随机切成224x224 大小图片 统一图片格式 transforms.RandomResizedCrop(224), # 图像翻转 transforms.RandomHorizontalFlip(), # totensor 归一化(0,255) >> (0,1) normalize channel=(channel-mean)/std transforms.ToTensor(), transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]) ]), "val" : transforms.Compose([ # 图片大小缩放 统一图片格式 transforms.Resize(256), # 以中心裁剪 transforms.CenterCrop(224), transforms.ToTensor(), transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]) ])}

导入,加载数据:

data_dir = './hymenoptera_data'# trans dataimage_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x), data_transforms[x]) for x in ['train', 'val']}# load datadata_loaders = {x: DataLoader(image_datasets[x], batch_size=BATCH_SIZE, shuffle=True) for x in ['train', 'val']}data_sizes = {x: len(image_datasets[x]) for x in ['train', 'val']}class_names = image_datasets['train'].classesprint(data_sizes, class_names)

{'train': 244, 'val': 153} ['ants', 'bees']

训练集 244图片 , 测试集153图片 。

可视化部分图片看看,由于visdom支持tensor输入 ,不用换成numpy,直接用tensor计算即可 :

inputs, classes = next(iter(data_loaders['val']))out = torchvision.utils.make_grid(inputs)inp = torch.transpose(out, 0, 2)mean = torch.FloatTensor([0.485, 0.456, 0.406])std = torch.FloatTensor([0.229, 0.224, 0.225])inp = std * inp + meaninp = torch.transpose(inp, 0, 2)viz.images(inp)

创建CNN

net 根据上一篇的处理cifar10的改了一下规格:

class CNN(nn.Module): def __init__(self, in_dim, n_class): super(CNN, self).__init__() self.cnn = nn.Sequential( nn.BatchNorm2d(in_dim), nn.ReLU(True), nn.Conv2d(in_dim, 16, 7), # 224 >> 218 nn.BatchNorm2d(16), nn.ReLU(inplace=True), nn.MaxPool2d(2, 2), # 218 >> 109 nn.ReLU(True), nn.Conv2d(16, 32, 5), # 105 nn.BatchNorm2d(32), nn.ReLU(True), nn.Conv2d(32, 64, 5), # 101 nn.BatchNorm2d(64), nn.ReLU(True), nn.Conv2d(64, 64, 3, 1, 1), nn.BatchNorm2d(64), nn.ReLU(True), nn.MaxPool2d(2, 2), # 101 >> 50 nn.Conv2d(64, 128, 3, 1, 1), # nn.BatchNorm2d(128), nn.ReLU(True), nn.MaxPool2d(3), # 50 >> 16 ) self.fc = nn.Sequential( nn.Linear(128*16*16, 120), nn.BatchNorm1d(120), nn.ReLU(True), nn.Linear(120, n_class)) def forward(self, x): out = self.cnn(x) out = self.fc(out.view(-1, 128*16*16)) return out# 输入3层rgb ,输出 分类 2 model = CNN(3, 2)

loss,优化函数:

line = viz.line(Y=np.arange(10))loss_f = nn.CrossEntropyLoss()optimizer = optim.SGD(model.parameters(), lr=LR, momentum=0.9)scheduler = optim.lr_scheduler.StepLR(optimizer, step_size=7, gamma=0.1)

参数:

BATCH_SIZE = 4LR = 0.001EPOCHS = 10

运行 10个 epoch 看看:

[9/10] train_loss:0.650|train_acc:0.639|test_loss:0.621|test_acc0.706
[10/10] train_loss:0.645|train_acc:0.627|test_loss:0.654|test_acc0.686
Training complete in 1m 16s
Best val Acc: 0.712418

运行 20个看看:

[19/20] train_loss:0.592|train_acc:0.701|test_loss:0.563|test_acc0.712
[20/20] train_loss:0.564|train_acc:0.721|test_loss:0.571|test_acc0.706
Training complete in 2m 30s
Best val Acc: 0.745098

准确率比较低:只有74.5%

我们使用models 里的 resnet18 运行 10个epoch:

model = torchvision.models.resnet18(True)num_ftrs = model.fc.in_featuresmodel.fc = nn.Linear(num_ftrs, 2)

[9/10] train_loss:0.621|train_acc:0.652|test_loss:0.588|test_acc0.667
[10/10] train_loss:0.610|train_acc:0.680|test_loss:0.561|test_acc0.667
Training complete in 1m 24s
Best val Acc: 0.686275

效果也很一般,想要短时间内就训练出效果很好的models,我们可以下载训练好的state,在此基础上训练:

model = torchvision.models.resnet18(pretrained=True)num_ftrs = model.fc.in_featuresmodel.fc = nn.Linear(num_ftrs, 2)

[9/10] train_loss:0.308|train_acc:0.877|test_loss:0.160|test_acc0.941
[10/10] train_loss:0.267|train_acc:0.885|test_loss:0.148|test_acc0.954
Training complete in 1m 25s
Best val Acc: 0.954248

10个epoch直接的到95%的准确率。

示例代码:https://github.com/ffzs/ml_pytorch/blob/master/ml_pytorch_hymenoptera

以上就是本文的全部内容,希望对大家的学习有所帮助,也希望大家多多支持。

声明:本页内容来源网络,仅供用户参考;我单位不保证亦不表示资料全面及准确无误,也不保证亦不表示这些资料为最新信息,如因任何原因,本网内容或者用户因倚赖本网内容造成任何损失或损害,我单位将不会负任何法律责任。如涉及版权问题,请提交至online#300.cn邮箱联系删除。

相关文章