TensorFlow实现CGAN

    条件GAN就是在GAN的基础上加入了一个条件y，在生成器和判别器中加入条件参与训练，这样训练出来的模型可以根据设置的条件生成想到的图，一般条件可以为label。CGAN的论文为：《Conditional Generative Adversarial Nets》。CGAN的结构图如下：

CGAN的实现只需要在GAN的基础上稍作修改即可，代码如下：

#coding=utf-8
import pickle
import tensorflow as tf
import numpy as np
import matplotlib.gridspec as gridspec
import os
import shutil
from scipy.misc import imsave


# 定义一个mnist数据集的类
class mnistReader():  
    def __init__(self,mnistPath,onehot=True):  
        self.mnistPath=mnistPath
        self.onehot=onehot  
        self.batch_index=0
        print ('read:',self.mnistPath)
        fo = open(self.mnistPath, 'rb')
        self.train_set,self.valid_set,self.test_set = pickle.load(fo,encoding='bytes')
        fo.close()        
        self.data_label_train=list(zip(self.train_set[0],self.train_set[1]))
        np.random.shuffle(self.data_label_train)               

    # 获取下一个训练集的batch
    def next_train_batch(self,batch_size=100):
        if self.batch_index < int(len(self.data_label_train)/batch_size):  
            # print ("batch_index:",self.batch_index )
            datum=self.data_label_train[self.batch_index*batch_size:(self.batch_index+1)*batch_size]  
            self.batch_index+=1  
            return self._decode(datum,self.onehot)  
        else:  
            self.batch_index=0  
            np.random.shuffle(self.data_label_train)  
            datum=self.data_label_train[self.batch_index*batch_size:(self.batch_index+1)*batch_size]  
            self.batch_index+=1  
            return self._decode(datum,self.onehot)          
    
    # 获取测试集的数据
    def test_data(self):
        tdata,tlabel=self.test_set
        data_label_test=list(zip(tdata,tlabel))
        return self._decode(data_label_test,self.onehot)    
    
    # 把一个batch的训练数据转换为可以放入模型训练的数据