生成对抗网

旧城等待， 2022-07-16 06:54 141阅读 0赞

https://github.com/hindupuravinash/the-gan-zoo

[生成式模型 & 生成对抗网络——资料梳理（专访资料 + 论文分类）][_ _ _]

[对抗样本与生成式对抗网络][Link 1]

**[生成对抗网络Generative Adversarial Nets][Generative Adversarial Nets]**

[ 生成式对抗网络GAN研究进展（一）][GAN]

[ 生成式对抗网络GAN研究进展（二）——原始GAN][GAN_GAN]

[生成式对抗网络GAN研究进展（三）——条件GAN][GAN_GAN 1]

[ 生成式对抗网络GAN研究进展（四）——Laplacian Pyramid of Adversarial Networks，LAPGAN][GAN_Laplacian Pyramid of Adversarial Networks_LAPGAN]

[生成式对抗网络GAN研究进展（五）——Deep Convolutional Generative Adversarial Nerworks，DCGAN][GAN_Deep Convolutional Generative Adversarial Nerworks_DCGAN]

[对抗生成网络（Generative Adversarial Net)][Generative Adversarial Net]

[深度学习（五十三）对抗网络][Link 2]

[DCGAN的小尝试（1）][DCGAN_1]

[在caffe 中实现Generative Adversarial Nets（一）][caffe _Generative Adversarial Nets]

[在caffe 中实现Generative Adversarial Nets（二）][caffe _Generative Adversarial Nets 1]

[提高驾驶技术：用GAN去除(爱情)动作片中的马赛克和衣服][GAN 1]

[GAN学习指南：从原理入门到制作生成Demo][GAN_Demo]

[手把手教你写一个生成对抗网络 — 生成对抗网络代码全解析, 详细代码解析(TensorFlow, numpy, matplotlib, scipy)][_ _ _TensorFlow_ numpy_ matplotlib_ scipy]

[基于能量模型的生成对抗网络–生成对抗网络进阶][Link 3]

[深度解读：GAN模型及其在2016年度的进展][GAN_2016]

[GAN: Generative Adversarial Nets][GAN_ Generative Adversarial Nets]

[近期GAN的模型和理论发展][GAN 2]

http://lijiancheng0614.github.io/all-categories/

http://blog.evjang.com/2016/06/generative-adversarial-nets-in.html

[Newmu/dcgan\_code][Newmu_dcgan_code]

[goodfeli/adversarial][goodfeli_adversarial]

https://plus.google.com/+SoumithChintala/posts/MCtDVqsef6f

http://blog.csdn.net/zhangjunhit/article/category/6869538

http://blog.csdn.net/AMDS123/article/category/6687276

[深度对抗学习在图像分割和超分辨率中的应用][Link 4]

[生成式对抗网络GAN有哪些最新的发展，可以实际应用到哪些场景中？][GAN 3]

[基于CNN的图像修复（CNN-based Image Inpainting）][CNN_CNN-based Image Inpainting]

[Image De-raining Using a Conditional Generative Adversarial Network][]

配套代码：https://github.com/hezhangsprinter/ID-CGAN

## WGAN ##

[看穿机器学习（W-GAN模型）的黑箱][W-GAN]

[Wasserstein GAN 的 TensorFlow 实现][Wasserstein GAN _ TensorFlow]

WGAN 模型的 pytorch 代码实现  
https://gist.github.com/soumith/71995cecc5b99cda38106ad64503cee3

[直观理解GAN背后的原理：以人脸图像生成为例][GAN 4]

[GAN用于半监督学习][GAN 5]

[SimGAN：活用合成数据和无监督数据][SimGAN]

[GAN之父NIPS 2016演讲现场直击：全方位解读生成对抗网络的原理及未来（附PPT）][GAN_NIPS 2016_PPT]

[GAN（Generative Adversarial Nets）研究进展][GAN_Generative Adversarial Nets]

[一文帮你发现各种出色的GAN变体][GAN 6]

[生成对抗网络新进展与论文全集][Link 5]

[生成对抗网络及其变体的论文汇总][Link 6]

[最小二乘GAN：比常规GAN更稳定，比WGAN收敛更迅速][GAN_GAN_WGAN]

[如何评价谷歌最近在人脸数据集上取得惊人效果的BEGAN模型？][BEGAN]

'''Deep Convolutional Generative Adversarial Network (DCGAN) Tutorial
    This tutorials walks through an implementation of DCGAN as described in Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks.
    To learn more about generative adversarial networks, see my Medium post on them.
    '''
    #Import the libraries we will need.
    import tensorflow as tf
    import numpy as np
    import input_data
    import matplotlib.pyplot as plt
    import tensorflow.contrib.slim as slim
    import os
    import scipy.misc
    import scipy
    '''We will be using the MNIST dataset. input_data is a library that downloads the dataset and uzips it automatically. It can be acquired Github here: https://gist.github.com/awjuliani/1d21151bc17362bf6738c3dc02f37906
    '''
    mnist = input_data.read_data_sets("MNIST_data/", one_hot=False)
    #Helper Functions
    
    #This function performns a leaky relu activation, which is needed for the discriminator network.
    def lrelu(x, leak=0.2, name="lrelu"):
         with tf.variable_scope(name):
             f1 = 0.5 * (1 + leak)
             f2 = 0.5 * (1 - leak)
             return f1 * x + f2 * abs(x)
        
    #The below functions are taken from carpdem20's implementation https://github.com/carpedm20/DCGAN-tensorflow
    #They allow for saving sample images from the generator to follow progress
    def save_images(images, size, image_path):
        return imsave(inverse_transform(images), size, image_path)
    
    def imsave(images, size, path):
        return scipy.misc.imsave(path, merge(images, size))
    
    def inverse_transform(images):
        return (images+1.)/2.
    
    def merge(images, size):
        h, w = images.shape[1], images.shape[2]
        img = np.zeros((h * size[0], w * size[1]))
    
        for idx, image in enumerate(images):
            i = idx % size[1]
            j = idx / size[1]
            img[j*h:j*h+h, i*w:i*w+w] = image
    
        return img
    '''Defining the Adversarial Networks
    Generator Network
    The generator takes a vector of random numbers and transforms it into a 32x32 image. Each layer in the network involves a strided transpose convolution, batch normalization, and rectified nonlinearity. Tensorflow's slim library allows us to easily define each of these layers.
    '''
    def generator(z):
        
        zP = slim.fully_connected(z,4*4*256,normalizer_fn=slim.batch_norm,\
            activation_fn=tf.nn.relu,scope='g_project',weights_initializer=initializer)
        zCon = tf.reshape(zP,[-1,4,4,256])
        
        gen1 = slim.convolution2d_transpose(\
            zCon,num_outputs=64,kernel_size=[5,5],stride=[2,2],\
            padding="SAME",normalizer_fn=slim.batch_norm,\
            activation_fn=tf.nn.relu,scope='g_conv1', weights_initializer=initializer)
        
        gen2 = slim.convolution2d_transpose(\
            gen1,num_outputs=32,kernel_size=[5,5],stride=[2,2],\
            padding="SAME",normalizer_fn=slim.batch_norm,\
            activation_fn=tf.nn.relu,scope='g_conv2', weights_initializer=initializer)
        
        gen3 = slim.convolution2d_transpose(\
            gen2,num_outputs=16,kernel_size=[5,5],stride=[2,2],\
            padding="SAME",normalizer_fn=slim.batch_norm,\
            activation_fn=tf.nn.relu,scope='g_conv3', weights_initializer=initializer)
        
        g_out = slim.convolution2d_transpose(\
            gen3,num_outputs=1,kernel_size=[32,32],padding="SAME",\
            biases_initializer=None,activation_fn=tf.nn.tanh,\
            scope='g_out', weights_initializer=initializer)
        
        return g_out
    '''Discriminator Network
    The discriminator network takes as input a 32x32 image and transforms it into a single valued probability of being generated from real-world data. Again we use tf.slim to define the convolutional layers, batch normalization, and weight initialization.
    '''
    def discriminator(bottom, reuse=False):
        
        dis1 = slim.convolution2d(bottom,16,[4,4],stride=[2,2],padding="SAME",\
            biases_initializer=None,activation_fn=lrelu,\
            reuse=reuse,scope='d_conv1',weights_initializer=initializer)
        
        dis2 = slim.convolution2d(dis1,32,[4,4],stride=[2,2],padding="SAME",\
            normalizer_fn=slim.batch_norm,activation_fn=lrelu,\
            reuse=reuse,scope='d_conv2', weights_initializer=initializer)
        
        dis3 = slim.convolution2d(dis2,64,[4,4],stride=[2,2],padding="SAME",\
            normalizer_fn=slim.batch_norm,activation_fn=lrelu,\
            reuse=reuse,scope='d_conv3',weights_initializer=initializer)
        
        d_out = slim.fully_connected(slim.flatten(dis3),1,activation_fn=tf.nn.sigmoid,\
            reuse=reuse,scope='d_out', weights_initializer=initializer)
        
        return d_out
    #Connecting them together
    
    tf.reset_default_graph()
    
    z_size = 100 #Size of z vector used for generator.
    
    #This initializaer is used to initialize all the weights of the network.
    initializer = tf.truncated_normal_initializer(stddev=0.02)
    
    #These two placeholders are used for input into the generator and discriminator, respectively.
    z_in = tf.placeholder(shape=[None,z_size],dtype=tf.float32) #Random vector
    real_in = tf.placeholder(shape=[None,32,32,1],dtype=tf.float32) #Real images
    
    Gz = generator(z_in) #Generates images from random z vectors
    Dx = discriminator(real_in) #Produces probabilities for real images
    Dg = discriminator(Gz,reuse=True) #Produces probabilities for generator images
    
    #These functions together define the optimization objective of the GAN.
    d_loss = -tf.reduce_mean(tf.log(Dx) + tf.log(1.-Dg)) #This optimizes the discriminator.
    g_loss = -tf.reduce_mean(tf.log(Dg)) #This optimizes the generator.
    
    tvars = tf.trainable_variables()
    
    #The below code is responsible for applying gradient descent to update the GAN.
    trainerD = tf.train.AdamOptimizer(learning_rate=0.0002,beta1=0.5)
    trainerG = tf.train.AdamOptimizer(learning_rate=0.0002,beta1=0.5)
    d_grads = trainerD.compute_gradients(d_loss,tvars[9:]) #Only update the weights for the discriminator network.
    g_grads = trainerG.compute_gradients(g_loss,tvars[0:9]) #Only update the weights for the generator network.
    
    update_D = trainerD.apply_gradients(d_grads)
    update_G = trainerG.apply_gradients(g_grads)
    '''Training the network
    Now that we have fully defined our network, it is time to train it!
    '''
    batch_size = 128 #Size of image batch to apply at each iteration.
    iterations = 500000 #Total number of iterations to use.
    sample_directory = './figs' #Directory to save sample images from generator in.
    model_directory = './models' #Directory to save trained model to.
    
    init = tf.initialize_all_variables()
    saver = tf.train.Saver()
    with tf.Session() as sess:  
        sess.run(init)
        for i in range(iterations):
            zs = np.random.uniform(-1.0,1.0,size=[batch_size,z_size]).astype(np.float32) #Generate a random z batch
            xs,_ = mnist.train.next_batch(batch_size) #Draw a sample batch from MNIST dataset.
            xs = (np.reshape(xs,[batch_size,28,28,1]) - 0.5) * 2.0 #Transform it to be between -1 and 1
            xs = np.lib.pad(xs, ((0,0),(2,2),(2,2),(0,0)),'constant', constant_values=(-1, -1)) #Pad the images so the are 32x32
            _,dLoss = sess.run([update_D,d_loss],feed_dict={z_in:zs,real_in:xs}) #Update the discriminator
            _,gLoss = sess.run([update_G,g_loss],feed_dict={z_in:zs}) #Update the generator, twice for good measure.
            _,gLoss = sess.run([update_G,g_loss],feed_dict={z_in:zs})
            if i % 10 == 0:
                print "Gen Loss: " + str(gLoss) + " Disc Loss: " + str(dLoss)
                z2 = np.random.uniform(-1.0,1.0,size=[batch_size,z_size]).astype(np.float32) #Generate another z batch
                newZ = sess.run(Gz,feed_dict={z_in:z2}) #Use new z to get sample images from generator.
                if not os.path.exists(sample_directory):
                    os.makedirs(sample_directory)
                #Save sample generator images for viewing training progress.
                save_images(np.reshape(newZ[0:36],[36,32,32]),[6,6],sample_directory+'/fig'+str(i)+'.png')
            if i % 1000 == 0 && i != 0:
                if not os.path.exists(model_directory):
                    os.makedirs(model_directory)
                saver.save(sess,model_directory+'/model-'+str(i)+'.cptk')
                print "Saved Model"
    '''Using a trained network
    Once we have a trained model saved, we may want to use it to generate new images, and explore the representation it has learned.
    '''
    sample_directory = './figs' #Directory to save sample images from generator in.
    model_directory = './models' #Directory to load trained model from.
    batch_size_sample = 36
    
    init = tf.initialize_all_variables()
    saver = tf.train.Saver()
    with tf.Session() as sess:  
        sess.run(init)
        #Reload the model.
        print 'Loading Model...'
        ckpt = tf.train.get_checkpoint_state(path)
        saver.restore(sess,ckpt.model_checkpoint_path)
        
        zs = np.random.uniform(-1.0,1.0,size=[batch_size_sample,z_size]).astype(np.float32) #Generate a random z batch
        newZ = sess.run(Gz,feed_dict={z_in:z2}) #Use new z to get sample images from generator.
        if not os.path.exists(sample_directory):
            os.makedirs(sample_directory)
        save_images(np.reshape(newZ[0:batch_size_sample],[36,32,32]),[6,6],sample_directory+'/fig'+str(i)+'.png')

[_ _ _]: http://blog.csdn.net/solomon1558/article/details/52338052
[Link 1]: http://blog.csdn.net/luoyun614/article/details/52202348
[Generative Adversarial Nets]: http://blog.csdn.net/shenxiaolu1984/article/details/52215983
[GAN]: http://blog.csdn.net/solomon1558/article/details/52537114
[GAN_GAN]: http://blog.csdn.net/solomon1558/article/details/52549409
[GAN_GAN 1]: http://blog.csdn.net/solomon1558/article/details/52555083
[GAN_Laplacian Pyramid of Adversarial Networks_LAPGAN]: http://blog.csdn.net/solomon1558/article/details/52562851
[GAN_Deep Convolutional Generative Adversarial Nerworks_DCGAN]: http://blog.csdn.net/solomon1558/article/details/52573596
[Generative Adversarial Net]: http://blog.csdn.net/stdcoutzyx/article/details/53151038
[Link 2]: http://blog.csdn.net/hjimce/article/details/57082684
[DCGAN_1]: https://zhuanlan.zhihu.com/p/22386494
[caffe _Generative Adversarial Nets]: http://blog.csdn.net/seven_first/article/details/52999758
[caffe _Generative Adversarial Nets 1]: http://blog.csdn.net/seven_first/article/details/53100325
[GAN 1]: http://www.cnblogs.com/frombeijingwithlove/p/7077448.html
[GAN_Demo]: http://geek.csdn.net/news/detail/134553
[_ _ _TensorFlow_ numpy_ matplotlib_ scipy]: http://nooverfit.com/wp/%E6%89%8B%E6%8A%8A%E6%89%8B%E6%95%99%E4%BD%A0%E5%86%99%E4%B8%80%E4%B8%AA%E7%94%9F%E6%88%90%E5%AF%B9%E6%8A%97%E7%BD%91%E7%BB%9C-%E7%94%9F%E6%88%90%E5%AF%B9%E6%8A%97%E7%BD%91%E7%BB%9C%E4%BB%A3/
[Link 3]: http://nooverfit.com/wp/
[GAN_2016]: http://www.leiphone.com/news/201701/Hrv1qUS6GYl1vl7O.html
[GAN_ Generative Adversarial Nets]: http://blog.csdn.net/helei001/article/details/70212099
[GAN 2]: http://mp.weixin.qq.com/s?__biz=MzI1NTE4NTUwOQ==&mid=2650325352&idx=1&sn=90fb15cee44fa7175a804418259d352e&scene=21#wechat_redirect
[Newmu_dcgan_code]: https://github.com/Newmu/dcgan_code
[goodfeli_adversarial]: https://github.com/goodfeli/adversarial
[Link 4]: https://zhuanlan.zhihu.com/p/25201511
[GAN 3]: https://www.zhihu.com/question/52602529/answer/163929347
[CNN_CNN-based Image Inpainting]: http://blog.csdn.net/yhl_leo/article/details/56674833
[Image De-raining Using a Conditional Generative Adversarial Network]: http://www.rci.rutgers.edu/~vmp93/index_ImageDeRaining.html
[W-GAN]: https://mp.weixin.qq.com/s?__biz=MzA3MzI4MjgzMw==&mid=2650723168&idx=3&sn=41fcf2fb0408c7b6a9b82d55d91c2b9c&chksm=871b171eb06c9e082c4083ff32748104a617e5cb1e6bd4d296b4db431358b8a41f40908ea8a5&scene=21#wechat_redirect
[Wasserstein GAN _ TensorFlow]: https://mp.weixin.qq.com/s?__biz=MzA3MzI4MjgzMw==&mid=2650723201&idx=3&sn=204b20981a52c8e190624e0c0e445857&chksm=871b17ffb06c9ee9b4fc57ea05762dde2f54229f7510c1f60deb7b6707ec3647066aa192a0a5&scene=21#wechat_redirect
[GAN 4]: https://mp.weixin.qq.com/s?__biz=MzA3MzI4MjgzMw==&mid=2650723168&idx=2&sn=68b21b815688443a0dd7caa115cc13fa&chksm=871b171eb06c9e085ab0f2223e6bab04d2eecfa430ca071b19d5c377d909d46724269126dbf3&scene=21#wechat_redirect
[GAN 5]: http://blog.csdn.net/shenxiaolu1984/article/details/75736407
[SimGAN]: http://blog.csdn.net/shenxiaolu1984/article/details/76160306
[GAN_NIPS 2016_PPT]: https://mp.weixin.qq.com/s?__biz=MzA3MzI4MjgzMw==&mid=2650721284&idx=1&sn=427e7f45c8253ab22a3960978409f5d1&chksm=871b087ab06c816c424ad03810be3e1b3aa9d6e99a5f325047796f110d178a07736f667d1a10&scene=21#wechat_redirect
[GAN_Generative Adversarial Nets]: https://mp.weixin.qq.com/s?__biz=MzA3MzI4MjgzMw==&mid=2650722118&idx=4&sn=92b16642541dfbc2a154e5c64bf2ffeb&chksm=871b0b38b06c822ea910fa2877a37effa5baf2b7c43107d927d70398a6f0ff7afed99fcf710d&scene=21#wechat_redirect
[GAN 6]: https://mp.weixin.qq.com/s?__biz=MzA3MzI4MjgzMw==&mid=2650724769&idx=2&sn=6fa540106cf6a5fd55fc39d057092888&chksm=871b1ddfb06c94c9e11d3a8281f60c0fce06a4e021fcd8eaab858c7f08ab9c939c4ad130e4b2&scene=21#wechat_redirect
[Link 5]: http://www.jiqizhixin.com/article/2671?utm_source=tuicool&utm_medium=referral
[Link 6]: http://www.jiqizhixin.com/article/index/id/2688
[GAN_GAN_WGAN]: https://mp.weixin.qq.com/s?__biz=MzA3MzI4MjgzMw==&mid=2650723893&idx=5&sn=bed0e585bba8ad01824a8d80603dbea6&chksm=871b124bb06c9b5d5f3b290ad02fb43ba02bdf7cbd04d53cfe839213ad7e57c0426555500e24&scene=21#wechat_redirect
[BEGAN]: https://www.zhihu.com/question/58358032