生成对抗网络,是一种基于博弈思想的网络训练思路,其主要网络模块由两部分组成,分别为generator(生成器)和discriminator(判别器)。
我们以GAM生成Minist手写数据集为例,在这个例子中,我们的目的是为了生成可以以假乱真的手写数字图片。而我们的训练思路,是使用生成器来产生一张照片,并且由判别器来判断这张照片是否是真实的照片。
在这个过程中,生成器会根据判别器返回的结果,一步一步的学习,生成器学习的目的是产生可以骗过判别器的照片,而判别器在这个过程中也会不断的进行学习,其学习的目的是,可以正确的分别出假图片和真图片。
#生成网络
def generator(noise_dim=NOISE_DIM):
net = nn.Sequential(
nn.Linear(noise_dim, 1024),
nn.ReLU(True),
nn.Linear(1024, 1024),
nn.ReLU(True),
nn.Linear(1024, 784), #最终输出大小为784
nn.Tanh()
)
return net
# 判别网络
def discriminator():
net = nn.Sequential(
nn.Linear(784, 256),
nn.LeakyReLU(0.2),
nn.Linear(256, 256),
nn.LeakyReLU(0.2),
nn.Linear(256, 1) #输出结果为置信度
)
return net
按照生成对抗网络的基本原理,我们对生成器和判别器的损失函数进行定义。
real_data = Variable(x).view(bs, -1) # 真实数据
logits_real = D_net(real_data) # 判别网络得分
sample_noise = (torch.rand(bs, noise_size) - 0.5) / 0.5 # -1 ~ 1 的均匀分布
g_fake_seed = Variable(sample_noise)
fake_images = G_net(g_fake_seed) # 生成的假的数据
logits_fake = D_net(fake_images) # 判别网络得分
d_total_error = discriminator_loss(logits_real, logits_fake) # 判别器的 loss
def discriminator_loss(logits_real, logits_fake): # 判别器的 loss
size = logits_real.shape[0]
true_labels = Variable(torch.ones(size, 1)).float()
false_labels = Variable(torch.zeros(size, 1)).float()
loss = bce_loss(logits_real, true_labels) + bce_loss(logits_fake, false_labels)
return loss
# 生成网络
g_fake_seed = Variable(sample_noise)
fake_images = G_net(g_fake_seed) # 生成的假的数据
gen_logits_fake = D_net(fake_images)
g_error = generator_loss(gen_logits_fake) # 生成网络的 loss
def generator_loss(logits_fake): # 生成器的 loss
size = logits_fake.shape[0]
true_labels = Variable(torch.ones(size, 1)).float()
loss = bce_loss(logits_fake, true_labels)
return loss
import torch
from torch import nn
from torch.autograd import Variable
import torchvision.transforms as tfs
from torch.utils.data import DataLoader, sampler
from torchvision.datasets import MNIST
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import os
plt.rcParams['figure.figsize'] = (10.0, 8.0) # 设置画图的尺寸
plt.rcParams['image.interpolation'] = 'nearest'
plt.rcParams['image.cmap'] = 'gray'
def show_images(images): # 定义画图工具
images = np.reshape(images, [images.shape[0], -1])
sqrtn = int(np.ceil(np.sqrt(images.shape[0])))
sqrtimg = int(np.ceil(np.sqrt(images.shape[1])))
fig = plt.figure(figsize=(sqrtn, sqrtn))
gs = gridspec.GridSpec(sqrtn, sqrtn)
gs.update(wspace=0.05, hspace=0.05)
for i, img in enumerate(images):
ax = plt.subplot(gs[i])
plt.axis('off')
ax.set_xticklabels([])
ax.set_yticklabels([])
ax.set_aspect('equal')
plt.imshow(img.reshape([sqrtimg, sqrtimg]))
return
def preprocess_img(x):
x = tfs.ToTensor()(x)
return (x - 0.5) / 0.5
def deprocess_img(x):
return (x + 1.0) / 2.0
class ChunkSampler(sampler.Sampler): # 定义一个取样的函数
"""Samples elements sequentially from some offset.
Arguments:
num_samples: # of desired datapoints
start: offset where we should start selecting from
"""
def __init__(self, num_samples, start=0):
self.num_samples = num_samples
self.start = start
def __iter__(self):
return iter(range(self.start, self.start + self.num_samples))
def __len__(self):
return self.num_samples
NUM_TRAIN = 50000
NUM_VAL = 5000
NOISE_DIM = 96
batch_size = 128
train_set = MNIST('./data', train=True, transform=preprocess_img,download=True)
train_data = DataLoader(train_set, batch_size=batch_size, sampler=ChunkSampler(NUM_TRAIN, 0))
val_set = MNIST('./data', train=True, transform=preprocess_img,download=True)
val_data = DataLoader(val_set, batch_size=batch_size, sampler=ChunkSampler(NUM_VAL, NUM_TRAIN))
imgs = deprocess_img(train_data.__iter__().next()[0].view(batch_size, 784)).numpy().squeeze() # 可视化图片效果
show_images(imgs)
# 判别网络
def discriminator():
net = nn.Sequential(
nn.Linear(784, 256),
nn.LeakyReLU(0.2),
nn.Linear(256, 256),
nn.LeakyReLU(0.2),
nn.Linear(256, 1)
)
return net
# 生成网络
def generator(noise_dim=NOISE_DIM):
net = nn.Sequential(
nn.Linear(noise_dim, 1024),
nn.ReLU(True),
nn.Linear(1024, 1024),
nn.ReLU(True),
nn.Linear(1024, 784),
nn.Tanh()
)
return net
# 判别器的 loss 就是将真实数据的得分判断为 1,假的数据的得分判断为 0,而生成器的 loss 就是将假的数据判断为 1
bce_loss = nn.BCEWithLogitsLoss() # 交叉熵损失函数
def discriminator_loss(logits_real, logits_fake): # 判别器的 loss
size = logits_real.shape[0]
true_labels = Variable(torch.ones(size, 1)).float()
false_labels = Variable(torch.zeros(size, 1)).float()
loss = bce_loss(logits_real, true_labels) + bce_loss(logits_fake, false_labels)
return loss
def generator_loss(logits_fake): # 生成器的 loss
size = logits_fake.shape[0]
true_labels = Variable(torch.ones(size, 1)).float()
loss = bce_loss(logits_fake, true_labels)
return loss
# 使用 adam 来进行训练,学习率是 3e-4, beta1 是 0.5, beta2 是 0.999
def get_optimizer(net):
optimizer = torch.optim.Adam(net.parameters(), lr=3e-4, betas=(0.5, 0.999))
return optimizer
def train_a_gan(D_net, G_net, D_optimizer, G_optimizer, discriminator_loss, generator_loss, show_every=250,
noise_size=96, num_epochs=10):
iter_count = 0
for epoch in range(num_epochs):
for x, _ in train_data:
bs = x.shape[0]
# 判别网络
real_data = Variable(x).view(bs, -1) # 真实数据
logits_real = D_net(real_data) # 判别网络得分
sample_noise = (torch.rand(bs, noise_size) - 0.5) / 0.5 # -1 ~ 1 的均匀分布
g_fake_seed = Variable(sample_noise)
fake_images = G_net(g_fake_seed) # 生成的假的数据
logits_fake = D_net(fake_images) # 判别网络得分
d_total_error = discriminator_loss(logits_real, logits_fake) # 判别器的 loss
D_optimizer.zero_grad()
d_total_error.backward()
D_optimizer.step() # 优化判别网络
# 生成网络
g_fake_seed = Variable(sample_noise)
fake_images = G_net(g_fake_seed) # 生成的假的数据
gen_logits_fake = D_net(fake_images)
g_error = generator_loss(gen_logits_fake) # 生成网络的 loss
G_optimizer.zero_grad()
g_error.backward()
G_optimizer.step() # 优化生成网络
if (iter_count % show_every == 0):
print('Iter: {}, D: {:.4}, G:{:.4}'.format(iter_count, d_total_error.item(), g_error.item()))
imgs_numpy = deprocess_img(fake_images.data.cpu().numpy())
show_images(imgs_numpy[0:16])
plt.show()
print()
iter_count += 1
D = discriminator()
G = generator()
D_optim = get_optimizer(D)
G_optim = get_optimizer(G)
train_a_gan(D, G, D_optim, G_optim, discriminator_loss, generator_loss)