pytorch cnn 识别手写的字实现自建图片数据_Python

本文主要介绍了pytorch cnn 识别手写的字实现自建图片数据，分享给大家，具体如下：
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
									# library

									# standard library

									import os 

									# third-party library

									import torch

									import torch.nn as nn

									from torch.autograd import Variable

									from torch.utils.data import Dataset, DataLoader

									import torchvision

									import matplotlib.pyplot as plt

									from PIL import Image

									import numpy as np

									# torch.manual_seed(1)  # reproducible 

									# Hyper Parameters

									EPOCH = 1        # train the training data n times, to save time, we just train 1 epoch

									BATCH_SIZE = 50

									LR = 0.001       # learning rate 

									root = "./mnist/raw/"

									def default_loader(path):

									  # return Image.open(path).convert('RGB')

									  return Image.open(path)

									class MyDataset(Dataset):

									  def __init__(self, txt, transform=None, target_transform=None, loader=default_loader):

									    fh = open(txt, 'r')

									    imgs = []

									    for line in fh:

									      line = line.strip('\n')

									      line = line.rstrip()

									      words = line.split()

									      imgs.append((words[0], int(words[1])))

									    self.imgs = imgs

									    self.transform = transform

									    self.target_transform = target_transform

									    self.loader = loader

									    fh.close()

									  def __getitem__(self, index):

									    fn, label = self.imgs[index]

									    img = self.loader(fn)

									    img = Image.fromarray(np.array(img), mode='L')

									    if self.transform is not None:

									      img = self.transform(img)

									    return img,label

									  def __len__(self):

									    return len(self.imgs)

									train_data = MyDataset(txt= root + 'train.txt', transform = torchvision.transforms.ToTensor())

									train_loader = DataLoader(dataset = train_data, batch_size=BATCH_SIZE, shuffle=True)

									test_data = MyDataset(txt= root + 'test.txt', transform = torchvision.transforms.ToTensor())

									test_loader = DataLoader(dataset = test_data, batch_size=BATCH_SIZE)

									class CNN(nn.Module):

									  def __init__(self):

									    super(CNN, self).__init__()

									    self.conv1 = nn.Sequential(     # input shape (1, 28, 28)

									      nn.Conv2d(

									        in_channels=1,       # input height

									        out_channels=16,      # n_filters

									        kernel_size=5,       # filter size

									        stride=1,          # filter movement/step

									        padding=2,         # if want same width and length of this image after con2d, padding=(kernel_size-1)/2 if stride=1

									      ),               # output shape (16, 28, 28)

									      nn.ReLU(),           # activation

									      nn.MaxPool2d(kernel_size=2),  # choose max value in 2x2 area, output shape (16, 14, 14)

									    )

									    self.conv2 = nn.Sequential(     # input shape (16, 14, 14)

									      nn.Conv2d(16, 32, 5, 1, 2),   # output shape (32, 14, 14)

									      nn.ReLU(),           # activation

									      nn.MaxPool2d(2),        # output shape (32, 7, 7)

									    )

									    self.out = nn.Linear(32 * 7 * 7, 10)  # fully connected layer, output 10 classes

									  def forward(self, x):

									    x = self.conv1(x)

									    x = self.conv2(x)

									    x = x.view(x.size(0), -1)      # flatten the output of conv2 to (batch_size, 32 * 7 * 7)

									    output = self.out(x)

									    return output, x  # return x for visualization 

									cnn = CNN()

									print(cnn) # net architecture

									optimizer = torch.optim.Adam(cnn.parameters(), lr=LR)  # optimize all cnn parameters

									loss_func = nn.CrossEntropyLoss()            # the target label is not one-hotted 

									# training and testing

									for epoch in range(EPOCH):

									  for step, (x, y) in enumerate(train_loader):  # gives batch data, normalize x when iterate train_loader

									    b_x = Variable(x)  # batch x

									    b_y = Variable(y)  # batch y

									    output = cnn(b_x)[0]        # cnn output

									    loss = loss_func(output, b_y)  # cross entropy loss

									    optimizer.zero_grad()      # clear gradients for this training step

									    loss.backward()         # backpropagation, compute gradients

									    optimizer.step()        # apply gradients

									    if step % 50 == 0:

									      cnn.eval()

									      eval_loss = 0.

									      eval_acc = 0.

									      for i, (tx, ty) in enumerate(test_loader):

									        t_x = Variable(tx)

									        t_y = Variable(ty)

									        output = cnn(t_x)[0]

									        loss = loss_func(output, t_y)

									        eval_loss += loss.data[0]

									        pred = torch.max(output, 1)[1]

									        num_correct = (pred == t_y).sum()

									        eval_acc += float(num_correct.data[0])

									      acc_rate = eval_acc / float(len(test_data))

									      print('Test Loss: {:.6f}, Acc: {:.6f}'.format(eval_loss / (len(test_data)), acc_rate))