700字范文 > 生死看淡不服就GAN（五）----用DCGAN生成MNIST手写体

生死看淡不服就GAN（五）----用DCGAN生成MNIST手写体

时间：2018-06-14 22:55:09

搭建DCGAN网络

#*************************************** 生死看淡，不服就GAN **************************************************************"""PROJECT:MNIST_DCGANAuthor:EphemeropteraDate:-4-25QQ:605686962Reference:' improved_wgan_training-master': </igul222/improved_wgan_training>'Zardinality/WGAN-tensorflow':</Zardinality/WGAN-tensorflow>'NELSONZHAO/zhihu':</NELSONZHAO/zhihu>""""""Note: in this section , we add batch-normalization-laysers in G\D to acclerate training.Additionally,we use moving average model to G to get well products from G"""# import dependencyimport tensorflow as tfimport numpy as npimport pickleimport visualizationimport osimport matplotlib.pyplot as pltfrom matplotlib.animation import FuncAnimationfrom threading import Threadimport timefrom time import sleepimport cv2# import MNIST datasetmnist_dir = r'../mnist_dataset'from tensorflow.examples.tutorials.mnist import input_datamnist = input_data.read_data_sets(mnist_dir)#--------------------------------------- define moudle -----------------------------------------------------------------# deconvdef deconv(img,new_size,fmaps,name='deconv'):with tf.variable_scope(name):img = tf.image.resize_nearest_neighbor(img,new_size,name='upscale')# upscalereturn tf.layers.conv2d(img,fmaps,3,padding='SAME',name='conv2d')# define Generatordef Generator_DC_28x28(latents,is_train):with tf.variable_scope("generator",reuse=(not is_train)):# dense0 ,size = (4,4), fmaps = 512dense0 = tf.layers.dense(latents,4*4*512,name='dense0')dense0 = tf.reshape(dense0,[-1,4,4,512])dense0 = tf.layers.batch_normalization(dense0, training=is_train)dense0 = tf.nn.leaky_relu(dense0)dense0 = tf.layers.dropout(dense0,rate=0.2)a = tf.get_variable_scope().name# deconv0 , size = (7,7) , fmaps = 256deconv1 = deconv(dense0,(7,7),256,name='deconv1')deconv1 = tf.layers.batch_normalization(deconv1, training=is_train)deconv1 = tf.nn.leaky_relu(deconv1)deconv1 = tf.layers.dropout(deconv1,rate=0.2)# deconv1 , size = (14,14) , fmaps = 128deconv2 = deconv(deconv1, (14, 14), 128, name='deconv2')deconv2 = tf.layers.batch_normalization(deconv2, training=is_train)deconv2 = tf.nn.leaky_relu(deconv2)deconv2 = tf.layers.dropout(deconv2, rate=0.2)# deconv2 , size = (28,28) , fmaps = 64deconv3 = deconv(deconv2, (28, 28), 64, name='deconv3')deconv3 = tf.layers.batch_normalization(deconv3, training=is_train)deconv3 = tf.nn.leaky_relu(deconv3)deconv3 = tf.layers.dropout(deconv3, rate=0.2)# toimg , size = (28,28) , fmaps = 1toimg = tf.layers.conv2d(deconv3,1,3,padding='SAME',bias_initializer=tf.zeros_initializer,activation=tf.nn.tanh,name='toimg')return toimg# define Discriminatordef Discriminator_DC_28x28(img,reuse = False):with tf.variable_scope("discriminator", reuse=reuse):# conv0 , size=(14,14) , fmaps =128conv0 = tf.layers.conv2d(img,128,3,padding='SAME',activation=tf.nn.leaky_relu,kernel_initializer=tf.random_normal_initializer(0,1), name='conv0')conv0 = tf.layers.average_pooling2d(conv0,2,2,padding='SAME',name='pool0')# conv1 , size=(7,7) , fmaps =256conv1 = tf.layers.conv2d(conv0, 256, 3, padding='SAME',kernel_initializer=tf.random_normal_initializer(0, 1), name='conv1')conv1 = tf.layers.batch_normalization(conv1,training=True)conv1 = tf.nn.leaky_relu(conv1)conv1 = tf.layers.average_pooling2d(conv1, 2, 2, padding='SAME', name='pool1')# conv2 , size=(5,5) , fmaps =512conv2 = tf.layers.conv2d(conv1, 512, 3, padding='VALID',kernel_initializer=tf.random_normal_initializer(0, 1), name='conv2')conv2 = tf.layers.batch_normalization(conv2, training=True)conv2 = tf.nn.leaky_relu(conv2)# dense3 ,size = 5*5*512dense3 = tf.reshape(conv2,[-1,5*5*512])dense3 = tf.layers.dense(dense3,1,name='dense3')outputs = tf.nn.sigmoid(dense3)return dense3,outputs# counting total to varsdef COUNT_VARS(vars):total_para = 0for variable in vars:# get each shape of varsshape = variable.get_shape()variable_para = 1for dim in shape:variable_para *= dim.valuetotal_para += variable_parareturn total_para# display paras infomationdef ShowParasList(paras):p = open('./trainLog/Paras.txt', 'w')p.writelines(['vars_total: %d'%COUNT_VARS(paras),'\n'])for variable in paras:p.writelines([variable.name, str(variable.get_shape()),'\n'])print(variable.name, variable.get_shape())p.close()# build related dirsdef GEN_DIR():if not os.path.isdir('ckpt'):print('DIR:ckpt NOT FOUND，BUILDING ON CURRENT PATH..')os.mkdir('ckpt')if not os.path.isdir('trainLog'):print('DIR:ckpt NOT FOUND，BUILDING ON CURRENT PATH..')os.mkdir('trainLog')#------------------------------------------ define grath ---------------------------------------------------------------# hyper-parameterslatents_dim = 128smooth = 0.1# define inputslatents = tf.placeholder(shape=[None,latents_dim],dtype=tf.float32,name='latents')input_real = tf.placeholder(shape=[None,28,28,1],dtype=tf.float32,name='input_real')# outputs from G\D# from Generatorg_outputs = Generator_DC_28x28(latents,is_train=True)g_test = Generator_DC_28x28(latents,is_train=False)# from Discriminatord_logits_real, d_outputs_real = Discriminator_DC_28x28(input_real,reuse=False)d_logits_fake, d_outputs_fake = Discriminator_DC_28x28(g_outputs,reuse=True)# define lossd_loss_real = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real,labels=tf.ones_like(d_logits_real)) * (1 - smooth))d_loss_fake = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake,labels=tf.zeros_like(d_logits_fake)))d_loss = tf.add(d_loss_real, d_loss_fake)g_loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake,labels=tf.ones_like(d_logits_fake)) * (1 - smooth))#-------------------------------------------- Paras Display ------------------------------------------------------------# list trainable varilablestrain_vars = tf.trainable_variables()# seperate into d/gd_train_vars = [var for var in train_vars if var.name.startswith("discriminator")]g_train_vars = [var for var in train_vars if var.name.startswith("generator")]# add g_train_vars to 'G_RAW'for var in g_train_vars:tf.add_to_collection('G_RAW',var)# list all varilablesall_vars = tf.all_variables()# get all vars of Gg_all_vars = [var for var in all_vars if var.name.startswith("generator")]# get μ，σ from BN of Gg_bn_m_v = [var for var in g_all_vars if 'moving_mean' in var.name]g_bn_m_v += [var for var in g_all_vars if 'moving_variance' in var.name]# add to 'G_BN_MV'for var in g_bn_m_v:tf.add_to_collection('G_BN_MV',var)#--------------------------------------------- Gradient Descent -------------------------------------------------------# training parameterslearn_rate = 2e-4G_step = tf.Variable(0, trainable=False)D_step = tf.Variable(0, trainable=False)# Gradient Descentwith tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)): # updating mean and std of batch firstd_train_opt = tf.train.AdamOptimizer(learn_rate,beta1=0.5).minimize(d_loss, var_list=d_train_vars,global_step=D_step)g_train_opt = tf.train.AdamOptimizer(learn_rate,beta1=0.5).minimize(g_loss, var_list=g_train_vars,global_step=G_step)#---------------------------------------- Exponential Moving Average for G ---------------------------------------------# apply EMAG_averages = tf.train.ExponentialMovingAverage(0.999, G_step)gvars_averages_op = G_averages.apply(g_train_vars) # apply ema# get shadowg_vars_ema = [G_averages.average(var) for var in g_train_vars] # g_train_vars using ema# add g_vars_ema to 'G_EMA'for ema in g_vars_ema:tf.add_to_collection('G_EMA',ema)# confirm training first and collect paras of G into [tf.GraphKeys.MOVING_AVERAGE_VARIABLES] laterwith tf.control_dependencies([g_train_opt,gvars_averages_op]):g_train_opt_ema = tf.no_op(name='g_train_opt_ema')#---------------------------------------------- iteration --------------------------------------------------------------# settingmax_iters = 5000batch_size = 50critic_n = 1# for recordingGEN_DIR()GenLog = []Losses = []saver = tf.train.Saver(var_list=g_train_vars+g_vars_ema+g_bn_m_v) # saving raw and ema# recording training infodef SavingRecords():global Lossesglobal GenLog# saving Losseswith open('./trainLog/loss_variation.loss', 'wb') as l:losses = np.array(Losses)pickle.dump(losses, l)print('saving Losses sucessfully！')# saving genlogwith open('./trainLog/GenLog.log', 'wb') as g:GenLog = np.array(GenLog)pickle.dump(GenLog, g)print('saving GenLog sucessfully！')# define trainingdef training():# runwith tf.Session() as sess:# initinit = (tf.global_variables_initializer(), tf.local_variables_initializer())sess.run(init)# Gotime_start = time.time() # gofor steps in range(max_iters+1):# get batchdata_batch = mnist.train.next_batch(batch_size)[0]# format modificationdata_batch = np.reshape(data_batch,[-1,28,28,1])# visualization.CV2_BATCH_RANDOM_SHOW(data_batch,1,25,5,5,0)data_batch = data_batch * 2 - 1data_batch = data_batch.astype(np.float32)# get latentsz = np.random.normal(0, 1, size=[batch_size, latents_dim]).astype(np.float32)# training discriminatorfor n in range(critic_n):sess.run(d_train_opt, feed_dict={input_real: data_batch, latents: z})# training generatorsess.run(g_train_opt_ema, feed_dict={input_real: data_batch,latents: z})# recording training_lossestrain_loss_d = sess.run(d_loss, feed_dict={input_real: data_batch, latents: z})train_loss_g = sess.run(g_loss, feed_dict={latents: z})info = [steps, train_loss_d, train_loss_g]# recording training_productsgen_sanmpes = sess.run(g_outputs, feed_dict={latents: z})visualization.CV2_BATCH_SHOW((gen_sanmpes[0:9] + 1) / 2, 0.5, 3, 3, delay=1)print('iters:：%d/%d..Discriminator_loss:%.3f..Generator_loss:%.3f..' % (steps, max_iters, train_loss_d, train_loss_g))if steps % 5 == 0:Losses.append(info)GenLog.append(gen_sanmpes)if steps % 1000 == 0 and steps > 0:saver.save(sess, './ckpt/generator.ckpt', global_step=steps)if steps == max_iters:# cv2.destroyAllWindows()# setup a thread to saving the training infotime_over = time.time()print('iterating is over! consuming time :%.2f'%(time_over-time_start))sleep(3)thread1 = Thread(target=SavingRecords, args=())thread1.start()yield info#------------------------------------------------ ANIMATION ------------------------------------------------------------# ANIMATION"""note: in this code , we will see the runtime-variation of G,D losses"""iters = []dloss = []gloss = []fig = plt.figure('runtime-losses')ax1 = fig.add_subplot(2,1,1,xlim=(0, max_iters), ylim=(-10, 10))ax2 = fig.add_subplot(2,1,2,xlim=(0, max_iters), ylim=(-20, 20))ax1.set_title('discriminator_loss')ax2.set_title('generator_loss')line1, = ax1.plot([], [], color='red',lw=1,label='discriminator')line2, = ax2.plot([], [],color='blue', lw=1,label='generator')fig.tight_layout()def init():line1.set_data([], [])line2.set_data([], [])return line1,line2def update(info):iters.append(info[0])dloss.append(info[1])gloss.append(info[2])line1.set_data(iters, dloss)line2.set_data(iters, gloss)return line1, line2ani = FuncAnimation(fig, update, frames=training,init_func=init, blit=True,interval=1,repeat=False)plt.show()