GAN

1. GAN(Generative Adversarial Network)

一个生成对抗网络包含两个均由多层神经网络构成的模型:(Generaive Model & Discriminative Model).G 产生仿真数据分布,D 判别数据是仿真还是真实的.GAN的训练过程是使G产生的仿真数据尽可能逼近真实数据,同时又使D尽量好地区分仿真数据和真实数据,目标是使G产生的数据足以以真乱假,是D判别真假的概率均为0.5.

#+BEGIN_SRC python output:results
import torch
import torch.nn as nn
import numpy as np
import matplotlib.pyplot as plt

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

# show our beautiful painting range
# 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.legend(loc='upper right')
# plt.show()


def artist_works():     # painting from the famous artist (real target)
    a = np.random.uniform(1, 2, size=BATCH_SIZE)[:, np.newaxis]
    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)

    prob_artist0 = D(artist_paintings)          # D try to increase this prob
    prob_artist1 = D(G_paintings)               # D try to reduce this prob

    D_loss = - torch.mean(torch.log(prob_artist0) + torch.log(1. - prob_artist1))
    G_loss = torch.mean(torch.log(1. - 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()

    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()

2. Cycle GAN

CycleGAN的创新点在于能够在源域和目标域之间，无须建立训练数据间一对一的映射，就可实现这种迁移.

想要做到这点，有两个比较重要的点，第一个就是双判别器。如下图所示，两个分布X,Y，生成器G，F分别是X到Y和Y到X的映射，两个判别器Dx,Dy可以对转换后的图片进行判别。第二个点就是cycle-consistency loss，用数据集中其他的图来检验生成器，这是防止G和F过拟合，比如想把一个小狗照片转化成梵高风格，如果没有cycle-consistency loss，生成器可能会生成一张梵高真实画作来骗过Dx，而无视输入的小狗。

Cycle Consistency 损失

D_A_loss_1 = tf.reduce_mean(tf.squared_difference(dec_A,1))
D_B_loss_1 = tf.reduce_mean(tf.squared_difference(dec_B,1))

D_A_loss_2 = tf.reduce_mean(tf.square(dec_gen_A))
D_B_loss_2 = tf.reduce_mean(tf.square(dec_gen_B))


D_A_loss = (D_A_loss_1 + D_A_loss_2)/2 #前向指导
D_B_loss = (D_B_loss_1 + D_B_loss_2)/2


g_loss_B_1 = tf.reduce_mean(tf.squared_difference(dec_gen_A,1))#反向生成
g_loss_A_1 = tf.reduce_mean(tf.squared_difference(dec_gen_A,1))


cyc_loss = tf.reduce_mean(tf.abs(input_A-cyc_A)) + tf.reduce_mean(tf.abs(input_B-cyc_B))
g_loss_A = g_loss_A_1 + 10*cyc_loss
g_loss_B = g_loss_B_1 + 10*cyc_loss

3. 接下来的工作

利用GAN预训练模型制作姿态估计数据集,自动生成模型.

对OpenCV考察调研

1. GAN(Generative Adversarial Network)

2. Cycle GAN

3. 接下来的工作

服务器配置