昇思25天学习打卡营第22天 | DCGAN生成漫画头像

DCGAN模型

深度卷积对抗生成网络（Depp Convolutional Generative Adversarial Networks, DCGAN）是GAN的直接拓展。区别在于DCGAN使用卷积和反卷积。

• 判别器：由分层的卷积层、BatchNorm层和LeakyReLU激活层组成。输入是$3\times 64\times 64$的图像，输出是该图像为真图像的概率。
• 生成器：由反卷积层、BatchNorm层和ReLU激活层组成，输入是标准正态分布中提取出的隐向量$z$，输出是$3\times 64\times 64$的RGB图像。

数据集

实验使用动漫头像数据集，共有70,171张动漫头像图片，大小均为$96\times 96$。

数据下载和超参数

from download import downloadurl = "https://download.mindspore.cn/dataset/Faces/faces.zip"path = download(url, "./faces", kind="zip", replace=True)batch_size = 128          # 批量大小
image_size = 64           # 训练图像空间大小
nc = 3                    # 图像彩色通道数
nz = 100                  # 隐向量的长度
ngf = 64                  # 特征图在生成器中的大小
ndf = 64                  # 特征图在判别器中的大小
num_epochs = 3           # 训练周期数
lr = 0.0002               # 学习率
beta1 = 0.5               # Adam优化器的beta1超参数

创建数据集

import numpy as np
import mindspore.dataset as ds
import mindspore.dataset.vision as visiondef create_dataset_imagenet(dataset_path):"""数据加载"""dataset = ds.ImageFolderDataset(dataset_path,num_parallel_workers=4,shuffle=True,decode=True)# 数据增强操作transforms = [vision.Resize(image_size),vision.CenterCrop(image_size),vision.HWC2CHW(),lambda x: ((x / 255).astype("float32"))]# 数据映射操作dataset = dataset.project('image')dataset = dataset.map(transforms, 'image')# 批量操作dataset = dataset.batch(batch_size)return datasetdataset = create_dataset_imagenet('./faces')

数据集可视化

import matplotlib.pyplot as pltdef plot_data(data):# 可视化部分训练数据plt.figure(figsize=(10, 3), dpi=140)for i, image in enumerate(data[0][:30], 1):plt.subplot(3, 10, i)plt.axis("off")plt.imshow(image.transpose(1, 2, 0))plt.show()sample_data = next(dataset.create_tuple_iterator(output_numpy=True))
plot_data(sample_data)

搭建网络

生成器

生成器$G$是将隐向量$z$映射到数据空间，由一系列Conv2dTranspose、BatchNorm2d和ReLU构成，输出数据经过tanh函数，使得返回$[-1,1]$范围的数据。

import mindspore as ms
from mindspore import nn, ops
from mindspore.common.initializer import Normalweight_init = Normal(mean=0, sigma=0.02)
gamma_init = Normal(mean=1, sigma=0.02)class Generator(nn.Cell):"""DCGAN网络生成器"""def __init__(self):super(Generator, self).__init__()self.generator = nn.SequentialCell(nn.Conv2dTranspose(nz, ngf * 8, 4, 1, 'valid', weight_init=weight_init),nn.BatchNorm2d(ngf * 8, gamma_init=gamma_init),nn.ReLU(),nn.Conv2dTranspose(ngf * 8, ngf * 4, 4, 2, 'pad', 1, weight_init=weight_init),nn.BatchNorm2d(ngf * 4, gamma_init=gamma_init),nn.ReLU(),nn.Conv2dTranspose(ngf * 4, ngf * 2, 4, 2, 'pad', 1, weight_init=weight_init),nn.BatchNorm2d(ngf * 2, gamma_init=gamma_init),nn.ReLU(),nn.Conv2dTranspose(ngf * 2, ngf, 4, 2, 'pad', 1, weight_init=weight_init),nn.BatchNorm2d(ngf, gamma_init=gamma_init),nn.ReLU(),nn.Conv2dTranspose(ngf, nc, 4, 2, 'pad', 1, weight_init=weight_init),nn.Tanh())def construct(self, x):return self.generator(x)generator = Generator()

判别器

判别器$D$是一个二分类网络，由Conv2d、BatchNorm2d和LeakyReLU构成，最后通过Sigmoid激活函数得到最终概率。

class Discriminator(nn.Cell):"""DCGAN网络判别器"""def __init__(self):super(Discriminator, self).__init__()self.discriminator = nn.SequentialCell(nn.Conv2d(nc, ndf, 4, 2, 'pad', 1, weight_init=weight_init),nn.LeakyReLU(0.2),nn.Conv2d(ndf, ndf * 2, 4, 2, 'pad', 1, weight_init=weight_init),nn.BatchNorm2d(ngf * 2, gamma_init=gamma_init),nn.LeakyReLU(0.2),nn.Conv2d(ndf * 2, ndf * 4, 4, 2, 'pad', 1, weight_init=weight_init),nn.BatchNorm2d(ngf * 4, gamma_init=gamma_init),nn.LeakyReLU(0.2),nn.Conv2d(ndf * 4, ndf * 8, 4, 2, 'pad', 1, weight_init=weight_init),nn.BatchNorm2d(ngf * 8, gamma_init=gamma_init),nn.LeakyReLU(0.2),nn.Conv2d(ndf * 8, 1, 4, 1, 'valid', weight_init=weight_init),)self.adv_layer = nn.Sigmoid()def construct(self, x):out = self.discriminator(x)out = out.reshape(out.shape[0], -1)return self.adv_layer(out)discriminator = Discriminator()

损失函数和优化器

使用二进制交叉熵损失函数BCELoss和Adam优化器：

# 定义损失函数
adversarial_loss = nn.BCELoss(reduction='mean')# 为生成器和判别器设置优化器
optimizer_D = nn.Adam(discriminator.trainable_params(), learning_rate=lr, beta1=beta1)
optimizer_G = nn.Adam(generator.trainable_params(), learning_rate=lr, beta1=beta1)
optimizer_G.update_parameters_name('optim_g.')
optimizer_D.update_parameters_name('optim_d.')

模型训练

• 判别器：最大化$\log D(x)+\log (1-D(G(z))$；
• 生成器：最小化$\log (1-D(G(z)))$。

def generator_forward(real_imgs, valid):# 将噪声采样为发生器的输入z = ops.standard_normal((real_imgs.shape[0], nz, 1, 1))# 生成一批图像gen_imgs = generator(z)# 损失衡量发生器绕过判别器的能力g_loss = adversarial_loss(discriminator(gen_imgs), valid)return g_loss, gen_imgsdef discriminator_forward(real_imgs, gen_imgs, valid, fake):# 衡量鉴别器从生成的样本中对真实样本进行分类的能力real_loss = adversarial_loss(discriminator(real_imgs), valid)fake_loss = adversarial_loss(discriminator(gen_imgs), fake)d_loss = (real_loss + fake_loss) / 2return d_lossgrad_generator_fn = ms.value_and_grad(generator_forward, None,optimizer_G.parameters,has_aux=True)
grad_discriminator_fn = ms.value_and_grad(discriminator_forward, None,optimizer_D.parameters)@ms.jit
def train_step(imgs):valid = ops.ones((imgs.shape[0], 1), mindspore.float32)fake = ops.zeros((imgs.shape[0], 1), mindspore.float32)(g_loss, gen_imgs), g_grads = grad_generator_fn(imgs, valid)optimizer_G(g_grads)d_loss, d_grads = grad_discriminator_fn(imgs, gen_imgs, valid, fake)optimizer_D(d_grads)return g_loss, d_loss, gen_imgsimport mindsporeG_losses = []
D_losses = []
image_list = []total = dataset.get_dataset_size()
for epoch in range(num_epochs):generator.set_train()discriminator.set_train()# 为每轮训练读入数据for i, (imgs, ) in enumerate(dataset.create_tuple_iterator()):g_loss, d_loss, gen_imgs = train_step(imgs)if i % 100 == 0 or i == total - 1:# 输出训练记录print('[%2d/%d][%3d/%d]   Loss_D:%7.4f  Loss_G:%7.4f' % (epoch + 1, num_epochs, i + 1, total, d_loss.asnumpy(), g_loss.asnumpy()))D_losses.append(d_loss.asnumpy())G_losses.append(g_loss.asnumpy())# 每个epoch结束后，使用生成器生成一组图片generator.set_train(False)fixed_noise = ops.standard_normal((batch_size, nz, 1, 1))img = generator(fixed_noise)image_list.append(img.transpose(0, 2, 3, 1).asnumpy())# 保存网络模型参数为ckpt文件mindspore.save_checkpoint(generator, "./generator.ckpt")mindspore.save_checkpoint(discriminator, "./discriminator.ckpt")

总结

这一节介绍了深度卷积生成对抗网络DCGAN，相对于经典的GAN网络来说，将生成器中的全连接层换成了反卷积层，而将判别器中的全连接层换成了卷积层，其训练过程和GAN网络基本一样。通过在70171张动漫头像上进行训练，使得该对抗网络能够生成动漫头像图片。

打卡