VGG16 finetuning实现猫狗分类
最近在学习VGG16网络的finetuning，想着在修改最后一层全连接层的基础上实现猫狗分类，但是在调试程序过程中发现了两个个好玩的东西：one—hot，权重部分载入。
关于one-hot的基础介绍请参考：点击打开链接
个人认为这个one-hot在一些条件下也可以不用，关键取决于你全连接输出时有没有对全连接层实行softmax，若进行了softmax操作，则必须进行one-hot，若无，则直接用标签即可，不用one-hot。
下面给出我进行finetuning进行猫狗大战的程序，用户只需更改文件地址即可训练，用户只需更改文件地址即可训练，用户只需更改文件地址即可训练，用户只需更改文件地址即可训练。重要的事情说四遍。
关于vgg16的权重文件下载请点击下面的链接：
权重下载链接
关于权重文件载入的一些说明：
因为我们要修改vgg16网络的最后一层全连接层，所以最后一层的权重我们不能载入，即最后两个部分（最后一层的权重跟偏置），实现代码为：
def load_weights(self, weight_file, sess):
weights = np.load(weight_file)
keys = sorted(weights.keys())
for i, k in enumerate(keys):
if i not in [30,31]:
print i, k, np.shape(weights[k])
sess.run(self.parameters[i].assign(weights[k]))
import tensorflow as tf
import numpy as np
import os
class vgg16:
def __init__(self, imgs, weights=None, sess=None):
self.imgs = imgs
self.convlayers()
self.fc_layers()
self.probs = tf.nn.softmax(self.fc3l)
if weights is not None and sess is not None:
self.load_weights(weights, sess)
def convlayers(self):
self.parameters = []
# zero-mean input
with tf.name_scope('preprocess') as scope:
mean = tf.constant([123.68, 116.779, 103.939], dtype=tf.float32, shape=[1, 1, 1, 3], name='img_mean')
images = self.imgs-mean
# conv1_1
with tf.name_scope('conv1_1') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 3, 64], dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
conv = tf.nn.conv2d(images, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[64], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv1_1 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# conv1_2
with tf.name_scope('conv1_2') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 64, 64], dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
conv = tf.nn.conv2d(self.conv1_1, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[64], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv1_2 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# pool1
self.pool1 = tf.nn.max_pool(self.conv1_2,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool1')
# conv2_1
with tf.name_scope('conv2_1') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 64, 128], dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
conv = tf.nn.conv2d(self.pool1, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[128], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv2_1 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# conv2_2
with tf.name_scope('conv2_2') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 128, 128], dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
conv = tf.nn.conv2d(self.conv2_1, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[128], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv2_2 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# pool2
self.pool2 = tf.nn.max_pool(self.conv2_2,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool2')
# conv3_1
with tf.name_scope('conv3_1') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 128, 256], dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
conv = tf.nn.conv2d(self.pool2, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[256], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv3_1 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# conv3_2
with tf.name_scope('conv3_2') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 256, 256], dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
conv = tf.nn.conv2d(self.conv3_1, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[256], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv3_2 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# conv3_3
with tf.name_scope('conv3_3') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 256, 256], dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
conv = tf.nn.conv2d(self.conv3_2, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[256], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv3_3 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# pool3
self.pool3 = tf.nn.max_pool(self.conv3_3,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool3')
# conv4_1
with tf.name_scope('conv4_1') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 256, 512], dtype=tf.float32,
stddev=1e-1), trainable=False,name='weights')
conv = tf.nn.conv2d(self.pool3, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[512], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv4_1 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# conv4_2
with tf.name_scope('conv4_2') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 512, 512], dtype=tf.float32,
stddev=1e-1), trainable=False,name='weights')
conv = tf.nn.conv2d(self.conv4_1, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[512], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv4_2 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# conv4_3
with tf.name_scope('conv4_3') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 512, 512], dtype=tf.float32,
stddev=1e-1), trainable=False,name='weights')
conv = tf.nn.conv2d(self.conv4_2, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[512], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv4_3 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# pool4
self.pool4 = tf.nn.max_pool(self.conv4_3,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool4')
# conv5_1
with tf.name_scope('conv5_1') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 512, 512], dtype=tf.float32,
stddev=1e-1), trainable=False,name='weights')
conv = tf.nn.conv2d(self.pool4, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[512], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv5_1 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# conv5_2
with tf.name_scope('conv5_2') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 512, 512], dtype=tf.float32,
stddev=1e-1), trainable=False,name='weights')
conv = tf.nn.conv2d(self.conv5_1, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[512], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv5_2 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# conv5_3
with tf.name_scope('conv5_3') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 512, 512], dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
conv = tf.nn.conv2d(self.conv5_2, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[512], dtype=tf.float32),
trainable=False, name='biases')
out = tf.nn.bias_add(conv, biases)
self.conv5_3 = tf.nn.relu(out, name=scope)
self.parameters += [kernel, biases]
# pool5
self.pool5 = tf.nn.max_pool(self.conv5_3,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME',
name='pool4')
def fc_layers(self):
# fc1
with tf.name_scope('fc1') as scope:
shape = int(np.prod(self.pool5.get_shape()[1:]))
fc1w = tf.Variable(tf.truncated_normal([shape, 4096],
dtype=tf.float32,
stddev=1e-1), trainable=False,name='weights')
fc1b = tf.Variable(tf.constant(1.0, shape=[4096], dtype=tf.float32),
trainable=False, name='biases')
pool5_flat = tf.reshape(self.pool5, [-1, shape])
fc1l = tf.nn.bias_add(tf.matmul(pool5_flat, fc1w), fc1b)
self.fc1 = tf.nn.relu(fc1l)
self.parameters += [fc1w, fc1b]
# fc2
with tf.name_scope('fc2') as scope:
fc2w = tf.Variable(tf.truncated_normal([4096, 4096],
dtype=tf.float32,
stddev=1e-1),trainable=False, name='weights')
fc2b = tf.Variable(tf.constant(1.0, shape=[4096], dtype=tf.float32),
trainable=False, name='biases')
fc2l = tf.nn.bias_add(tf.matmul(self.fc1, fc2w), fc2b)
self.fc2 = tf.nn.relu(fc2l)
self.parameters += [fc2w, fc2b]
# fc3
with tf.name_scope('fc3') as scope:
fc3w = tf.Variable(tf.truncated_normal([4096, 2],
dtype=tf.float32,
stddev=1e-1), trainable=True,name='weights')
fc3b = tf.Variable(tf.constant(1.0, shape=[2], dtype=tf.float32),
trainable=True, name='biases')
self.fc3l = tf.nn.bias_add(tf.matmul(self.fc2, fc3w), fc3b)
self.parameters += [fc3w, fc3b]
def load_weights(self, weight_file, sess):
weights = np.load(weight_file)
keys = sorted(weights.keys())
for i, k in enumerate(keys):
if i not in [30,31]:
print i, k, np.shape(weights[k])
sess.run(self.parameters[i].assign(weights[k]))
print("-------------all done-------------")
N_CLASSES = 2 # 2个输出神经元，［1，0］ 或者 ［0，1］猫和狗的概率
IMG_W = 224 # 重新定义图片的大小，图片如果过大则训练比较慢
IMG_H = 224
BATCH_SIZE = 32 #每批数据的大小
CAPACITY = 256 #队列容量,队列中最多容纳图片的个数
MAX_STEP = 15000 # 训练的步数，应当 >= 10000
learning_rate = 0.0001 # 学习率，建议刚开始的 learning_rate <= 0.0001
file_dir = 'catdog/train/'
def get_files(file_dir):
# file_dir: 文件夹路径
# return: 乱序后的图片和标签
cats = []
label_cats = []
dogs = []
label_dogs = []
# 载入数据路径并写入标签值
for file in os.listdir(file_dir):
name = file.split('.')
if name[0] == 'cat':
cats.append(file_dir + file) #修改原来的列表,增加对象
label_cats.append(0)
else:
dogs.append(file_dir + file)
label_dogs.append(1)
print("There are %d cats\nThere are %d dogs" % (len(cats), len(dogs)))
# 打乱文件顺序
image_list = np.hstack((cats, dogs)) #把猫放在前面，狗在后面组成一个文件
label_list = np.hstack((label_cats, label_dogs))
temp = np.array([image_list, label_list]) #生成序列
temp = temp.transpose() # 转置
np.random.shuffle(temp)
image_list = list(temp[:, 0]) #下标索引，访问列表值
label_list = list(temp[:, 1])
label_list = [int(i) for i in label_list]
return image_list, label_list
def get_batch(image,label,image_W,image_H,batch_size,capacity):
# 转换数据为 ts 能识别的格式
image = tf.cast(image,tf.string) #将image格式转化为字符串
label = tf.cast(label, tf.int32)
# 将image 和 label 放倒队列里
input_queue = tf.train.slice_input_producer([image,label])
label = input_queue[1]
# 读取图片的全部信息
image_contents = tf.read_file(input_queue[0])
# 把图片解码，channels ＝3 为彩色图片, r，g ，b 黑白图片为 1 ，也可以理解为图片的厚度
image = tf.image.decode_jpeg(image_contents,channels =3)
# 将图片以图片中心进行裁剪或者扩充为 指定的image_W，image_H
image = tf.image.resize_image_with_crop_or_pad(image, image_W, image_H)
# 对数据进行标准化,标准化，就是减去它的均值，除以他的方差
image = tf.image.per_image_standardization(image)
# 生成批次 num_threads 有多少个线程根据电脑配置设置 capacity 队列中 最多容纳图片的个数 tf.train.shuffle_batch 打乱顺序，
image_batch, label_batch = tf.train.batch([image, label],batch_size = batch_size, num_threads = 64, capacity = capacity)
# 重新定义下 label_batch 的形状
label_batch = tf.reshape(label_batch , [batch_size])
# 转化图片格式
image_batch = tf.cast(image_batch,tf.float32)
return image_batch, label_batch
def onehot(labels):
'''one-hot'''
n_sample = len(labels)
n_class = max(labels)+1
onehot_labels = np.zeros((n_sample ,n_class))
onehot_labels[np.arange(n_sample),labels] = 1
return onehot_labels
if __name__ == '__main__':
sess = tf.Session()
imgs = tf.placeholder(tf.float32, [None, 224, 224, 3])
y_=tf.placeholder(tf.float32,[None, 2])
vgg = vgg16(imgs, 'vgg16_weights.npz', sess)
#img1 = imread('2.jpg', mode='RGB')
#img1 = imresize(img1, (224, 224))
train_dir = 'catdog/train/' #My dir--20170727-csq
logs_train_dir = 'catdog/log/'
train, train_label = get_files(train_dir)
train_batch, train_label_batch = get_batch(train,
train_label,
IMG_W,
IMG_H,
BATCH_SIZE,
CAPACITY)
prob = vgg.probs
loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits \
(logits=prob, labels=y_, name='xentropy_per_example') )
optimizer = tf.train.AdamOptimizer(learning_rate= learning_rate).minimize(loss)
sess.run(tf.global_variables_initializer())
###TensorFlow提供了两个类来帮助多线程的实现：tf.Coordinator和 tf.QueueRunner。
#从设计上这两个类必须被一起使用。Coordinator类可以用来同时停止多个工作线程并且向那个在等待所有工作线程终止的程序报告异常。
#QueueRunner类用来协调多个工作线程同时将多个张量推入同一个队列中
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(sess=sess, coord=coord)
#print("ok")
for i in range(MAX_STEP):
image, label = sess.run([train_batch, train_label_batch])
label = onehot(label)
sess.run(optimizer,feed_dict = {imgs:image,y_:label})
loss_r = sess.run(loss,feed_dict = {imgs:image,y_:label})
if i % 5 == 0:
print('Step %d, train loss = %.2f'%(i, loss_r))
coord.join(threads)
sess.close()