GAN的简单思考及实例

参考资料：https://morvanzhou.github.io/tutorials/

详细说明已在程序注释给出。

"""
View more, visit my tutorial page: https://morvanzhou.github.io/tutorials/
My Youtube Channel: https://www.youtube.com/user/MorvanZhou

Dependencies:
torch: 1.1.0
numpy
matplotlib

修改作者：kindy
修改日期：2019-05-28

假设想要模仿一位大师的绘画作品（real），模仿的作品（Generator生成的fake样本）需要不断和大师的作品进行对比（Discriminator进行比较），当两者差距最小（Discriminator的目标函数criter function）时，说明模仿的作品足够以假乱真，这就是相互博弈的一个过程。

相关函数：
np.vstack(tup)：垂直方向拼接数据（first axis比较好理解），tup是一个tuple，例如a大小为(2,3)，
    b大小为(4,3)，那么通过:
>>> c = np.vstack((a,b))
    拼接之后的c大小为(6,3)，效果是数据a下面堆叠b（垂直方向）

np.random.uniform(a,b,size)：均匀分布U(a,b)

np.newaxis：增加维度，例如:
>>> a=rand(10)

>>> a.shape
>>> (10,)

>>> a[:,newaxis].shape
>>> (10, 1)

>>> a[newaxis,:].shape
>>> (1, 10)
"""
import torch
import torch.nn as nn
import numpy as np
import matplotlib.pyplot as plt
from torchvision import *
# torch.manual_seed(1)    # reproducible
# np.random.seed(1)

# Hyper Parameters
BATCH_SIZE = 64
LR_G = 0.0001           # learning rate for generator
LR_D = 0.0001           # learning rate for discriminator
N_IDEAS = 5             # think of this as number of ideas for generating an art work (Generator)
ART_COMPONENTS = 15     # it could be total point G can draw in the canvas，每个作品的维度

PAINT_POINTS = np.vstack([np.linspace(-1, 1, ART_COMPONENTS) for _ in range(BATCH_SIZE)]) # （64.15）

def artist_works():     # painting from the famous artist (real target)
    a = np.random.uniform(1, 2, size=BATCH_SIZE)[:, np.newaxis] # (64,1)
    paintings = a * np.power(PAINT_POINTS, 2) + (a-1)
    paintings = torch.from_numpy(paintings).float()
    return paintings

G = nn.Sequential(                      # Generator
    nn.Linear(N_IDEAS, 128),            # random ideas (could from normal distribution)
    nn.ReLU(),
    nn.Linear(128, ART_COMPONENTS),     # making a painting from these random ideas
)

D = nn.Sequential(                      # Discriminator
    nn.Linear(ART_COMPONENTS, 128),     # receive art work either from the famous artist or a newbie like G
    nn.ReLU(),
    nn.Linear(128, 1),
    nn.Sigmoid(),                       # tell the probability that the art work is made by artist
)

opt_D = torch.optim.Adam(D.parameters(), lr=LR_D)
opt_G = torch.optim.Adam(G.parameters(), lr=LR_G)

plt.ion()   # something about continuous plotting

for step in range(10000):
    artist_paintings = artist_works()           # real painting from artist
    G_ideas = torch.randn(BATCH_SIZE, N_IDEAS)  # random ideas
    G_paintings = G(G_ideas)                    # fake painting from G (random ideas)

    # D判断真画为真实画作的概率， try to increase this prob
    prob_artist0 = D(artist_paintings)
    # D判断赝品为真实画作的概率， try to reduce this prob
    prob_artist1 = D(G_paintings)

    # 下面的G_loss和D_loss是自定义的目标函数（损失函数），并非取自torch已有的损失函数

    # 对判别器而言，目标是能正确区分真画和赝品
    D_loss = - torch.mean(torch.log(prob_artist0) + torch.log(1. - prob_artist1))

    # 对生成器来说，肯定是模仿画判断为真画的概率越高越好，概率越高说明赝品越接近真品
    #所以我们的目标是增大 prob_artist1D(G_paintings) 的值，在torch里面优化目标是最小化，所以写成下面形式，可以单独绘制函数 f(x)=log(1-x)感受一下函数的形式
    # G_loss = torch.mean(torch.log(1. - prob_artist1))

    # 将 G_loss换成以下目标函数试试：
    G_loss = - torch.mean(torch.log(prob_artist1))

    opt_D.zero_grad()
    D_loss.backward(retain_graph=True)      # reusing computational graph
    opt_D.step()

    opt_G.zero_grad()
    G_loss.backward()
    opt_G.step()

    print('step:{}'.format(step))

    if step % 50 == 0:  # plotting
        plt.cla()
        plt.plot(PAINT_POINTS[0], G_paintings.data.numpy()[0], c='#4AD631', lw=3, label='Generated painting',)
        plt.plot(PAINT_POINTS[0], 2 * np.power(PAINT_POINTS[0], 2) + 1, c='#74BCFF', lw=3, label='upper bound')
        plt.plot(PAINT_POINTS[0], 1 * np.power(PAINT_POINTS[0], 2) + 0, c='#FF9359', lw=3, label='lower bound')
        plt.text(-.5, 2.3, 'D accuracy=%.2f (0.5 for D to converge)' % prob_artist0.data.numpy().mean(), fontdict={'size': 13})
        plt.text(-.5, 2, 'D score= %.2f (-1.38 for G to converge)' % -D_loss.data.numpy(), fontdict={'size': 13})
        plt.ylim((0, 3));plt.legend(loc='upper right', fontsize=10);plt.draw();plt.pause(0.01)

plt.ioff()
plt.show()

关键点：
（1）D，G的网络构建；
（2）D，G两个网络的目标函数选取问题，好多衍生的GAN都是在目标函数选取基础是实现的