VGGNet探索了卷积神经网络的深度与其性能之间的关系,通过反复堆叠3*3的小型卷积核和2*2的最大池化层,VGGNet成功地构筑了16~19层的卷积神经网络。VGGNet的结构非常简洁,整个网络都使用了同样大小的卷积核尺寸(3*3)和最大池化层(2*2)。
到目前为止,VGGNet依然经常被用来提取图像特征。VGGNet训练后的模型参数在其官网上开源了,可用来在domain specific的图像分类任务上进行再训练(相当于提供了非常好的初始化权重),因此被用在了很多地方。VGGNet的网络结构如下图所示:
在VGGNet中运用到的技巧:
- 经常出现多个完全一样的3*3的卷积层堆叠在一起的情况,这其实是非常有用的设计。例如2个3*3的卷积层串联相当于1个5*5的卷积层,即一个像素会跟周围5*5的像素产生关联,可以说感受野大小为5*5。而3个3*3的卷积层串联的效果相当于一个7*7的卷积层。但是3个3*3的卷积层拥有比1个7*7的卷积层更少的参数,只有后者的 3∗3∗37∗7=55 。最重要的是,3个3*3的卷积层拥有比一个7*7的卷积层更多的非线性变换,使得CNN对特征的学习能力更强。
- VGGNet在训练时,先训练级别A的简单网络,再复用A网络的权重来初始化后面的几个复杂模型,这样训练收敛的速度更快。
- 在预测时,VGG采用mutil-Scale的方法,将图像scale到一个尺寸Q,并将图像输入卷积网络计算。然后在最后一个卷积层使用滑窗的方式进行分类预测,将不同窗口的分类结果平均,再将不同尺寸Q的结果平均得到最后结果,这样可提高图片数据的利用率并提升预测准确率。
- VGGNet还使用了Mutil-Scale的方法做数据增强,将原始数据缩放到不同尺寸S,然后再随机裁切224*224的图片,这样能增加很多数据量,对于防止模型过拟合有不错的效果。
作者在对比各级网络时总结了一下观点:
- LRN层作用不大
- 越深的网络效果越好
- 1*1的卷积也是很有效的,但是没有3*3的卷积好,大一些的卷积核可以学习更大的空间特征。
在这里,我们不直接使用ImageNet数据训练一个VGGNet,而是采用跟AlexNet一样的方式:构造出VGGNet网络,并评测其forward(inference)耗时和backward(training)耗时。
from datetime import datetime
import math
import time
import tensorflow as tf
batch_size = 32
num_batches = 100
# 先定义一个conv_op函数,用于创建卷积层并把本层的参数存入参数列表
def conv_op(input_op, name, kh, kw, n_out, dh, dw, p):
n_in = input_op.get_shape()[-1].value
with tf.name_scope(name) as scope:
kernel = tf.get_variable(scope+"w",
shape=[kh, kw, n_in, n_out], dtype=tf.float32,
initializer=tf.contrib.layers.xavier_initializer_conv2d())
conv = tf.nn.conv2d(input_op, kernel, (1, dh, dw, 1), padding='SAME')
bias_init_val = tf.constant(0.0, shape=[n_out], dtype=tf.float32)
biases = tf.Variable(bias_init_val, trainable=True, name='b')
z = tf.nn.bias_add(conv, biases)
activation = tf.nn.relu(z, name=scope)
p += [kernel, biases]
return activation
# 下面定义全连接层创建函数fc_op
def fc_op(input_op, name, n_out, p):
n_in = input_op.get_shape()[-1].value
with tf.name_scope(name) as scope:
kernel = tf.get_variable(scope+"w",
shape=[n_in, n_out],
dtype=tf.float32,
initializer=tf.contrib.layers.xavier_initializer_conv2d())
biases = tf.Variable(tf.constant(0.1, shape=[n_out], dtype=tf.float32), name='b')
activation = tf.nn.relu_layer(input_op, kernel, biases, name=scope)
p += [kernel, biases]
return activation
# 定义最大池化层的创建函数mpool_op
def mpool_op(input_op, name, kh, kw, dh, dw):
return tf.nn.max_pool(input_op,
ksize=[1, kh, kw, 1],
strides=[1, dh, dw, 1],
padding='SAME',
name=name)
# 开始创建VGGNet-16,主要分为6个部分:前5部分为卷积网络,最后一段是全连接网络
def inference_op(input_op, keep_prob):
p = []
conv1_1 = conv_op(input_op, name='conv1_1', kh=3, kw=3, n_out=64, dh=1, dw=1, p=p)
conv1_2 = conv_op(conv1_1, name='conv1_2', kh=3, kw=3, n_out=64, dh=1, dw=1, p=p)
pool1 = mpool_op(conv1_2, name='pool1', kh=2, kw=2, dw=2, dh=2)
conv2_1 = conv_op(pool1, name='conv2_1', kh=3, kw=3, n_out=128, dh=1, dw=1, p=p)
conv2_2 = conv_op(conv2_1, name='conv2_2', kh=3, kw=3, n_out=128, dh=1, dw=1, p=p)
pool2 = mpool_op(conv2_2, name='pool2', kh=2, kw=2, dh=2, dw=2)
conv3_1 = conv_op(pool2, name='conv3_1', kh=3, kw=3, n_out=256, dh=1, dw=1, p=p)
conv3_2 = conv_op(conv3_1, name='conv3_2', kh=3, kw=3, n_out=256, dh=1, dw=1, p=p)
conv3_3 = conv_op(conv3_2, name='conv3_3', kh=3, kw=3, n_out=256, dh=1, dw=1, p=p)
pool3 = mpool_op(conv3_3, name='pool3', kh=2, kw=2, dh=2, dw=2)
conv4_1 = conv_op(pool3, name='conv4_1', kh=3, kw=3, n_out=512, dh=1, dw=1, p=p)
conv4_2 = conv_op(conv4_1, name='conv4_2', kh=3, kw=3, n_out=512, dh=1, dw=1, p=p)
conv4_3 = conv_op(conv4_2, name='conv4_3', kh=3, kw=3, n_out=512, dh=1, dw=1, p=p)
pool4 = mpool_op(conv4_3, name='pool4', kh=2, kw=2, dh=2, dw=2)
conv5_1 = conv_op(pool4, name='conv5_1', kh=3, kw=3, n_out=512, dh=1, dw=1, p=p)
conv5_2 = conv_op(conv5_1, name='conv5_2', kh=3, kw=3, n_out=512, dh=1, dw=1, p=p)
conv5_3 = conv_op(conv5_2, name='conv5_3', kh=3, kw=3, n_out=512, dh=1, dw=1, p=p)
pool5 = mpool_op(conv5_3, name='pool5', kh=2, kw=2, dh=2, dw=2)
# 将输出结果扁平化
shp = pool5.get_shape()
flattened_shape = shp[1].value * shp[2].value * shp[3].value
resh1 = tf.reshape(pool5, [-1, flattened_shape], name="resh1")
fc6 = fc_op(resh1, name='fc6', n_out=4096, p=p)
fc6_drop = tf.nn.dropout(fc6, keep_prob=keep_prob, name="fc6_drop")
fc7 = fc_op(fc6_drop, name='fc7', n_out=4096, p=p)
fc7_drop = tf.nn.dropout(fc7, keep_prob=keep_prob, name="fc7_drop")
fc8 = fc_op(fc7_drop, name='fc8', n_out=1000, p=p)
softmax = tf.nn.softmax(fc8)
predictions = tf.argmax(softmax, 1)
return predictions, softmax, fc8, p
# 定义测评函数time_tensorflow_run()
def time_tensorflow_run(session, target, feed, info_string):
num_steps_burn_in = 10
total_duration = 0.0
total_duration_squared = 0.0
for i in range(num_batches + num_steps_burn_in):
start_time = time.time()
_ = session.run(target, feed_dict=feed)
duration = time.time() - start_time
if i >= num_steps_burn_in:
if not i % 10:
print('%s: step %d, duration = %.3f' % (datetime.now(), i - num_steps_burn_in, duration))
total_duration += duration
total_duration_squared += duration * duration
mn = total_duration / num_batches
vr = total_duration_squared / num_batches - mn * mn
sd = math.sqrt(vr)
print('%s: %s across %d steps, %.3f +/- %.3f sec / batch' % (datetime.now(), info_string,
num_batches, mn, sd))
def run_benchmark():
with tf.Graph().as_default():
image_size = 224
images = tf.Variable(tf.random_normal([batch_size,
image_size,
image_size, 3],
dtype=tf.float32,
stddev=1e-1))
keep_prob = tf.placeholder(dtype=tf.float32)
predictions, softmax, fc8, p = inference_op(images, keep_prob)
init = tf.global_variables_initializer()
sess = tf.Session()
sess.run(init)
time_tensorflow_run(sess, predictions, {keep_prob: 1.0}, "Forward")
objective = tf.nn.l2_loss(fc8)
grad = tf.gradients(objective, p)
time_tensorflow_run(sess, grad, {keep_prob: 0.5}, "Forward-backward")
run_benchmark()
VGGNet的模型参数虽然比AlexNet多, 但反而只需要较少的迭代次数就可以收敛,主要原因是更深的网络和更小的卷积核带来的隐式的正则化效果。VGGNet凭借其相对不算很高的复杂度和优秀的分类性能,成为了一代经典的卷积神经网络,直到现在依然被应用在很多地方。