1. 池化和采样
-
Pooling
- Max/Avg pooling
- Max/Avg pooling
-
upsample
- nearest
- bilinear
-
Relu
2. cifar100
-
13 layers
-
Code
import tensorflow as tf
from tensorflow.keras import layers, optimizers, metrics, datasets, Sequential
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
tf.random.set_seed(2345)
conv_layers = [ # 5 units of conv + max pooling
# unit 1
layers.Conv2D(filters=64, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.Conv2D(filters=64, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.MaxPool2D(pool_size=[2, 2], strides=2, padding='same'),
# unit 2
layers.Conv2D(filters=128, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.Conv2D(filters=128, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.MaxPool2D(pool_size=[2, 2], strides=2, padding='same'),
# unit 3
layers.Conv2D(filters=256, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.Conv2D(filters=256, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.MaxPool2D(pool_size=[2, 2], strides=2, padding='same'),
# unit 4
layers.Conv2D(filters=512, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.Conv2D(filters=512, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.MaxPool2D(pool_size=[2, 2], strides=2, padding='same'),
# unit 5
layers.Conv2D(filters=512, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.Conv2D(filters=512, kernel_size=[3, 3], padding='same', activation=tf.nn.relu),
layers.MaxPool2D(pool_size=[2, 2], strides=2, padding='same'),
]
def pre_process(x, y):
# [0~1]
x = tf.cast(x, dtype=tf.float32) / 255.
y = tf.cast(y, dtype=tf.int32)
return x, y
# (50000, 32, 32, 3) (50000, 1) (10000, 32, 32, 3) (10000, 1)
(x, y), (x_test, y_test) = datasets.cifar100.load_data()
print(y_test)
y = tf.squeeze(y, axis=1)
train_db = tf.data.Dataset.from_tensor_slices((x, y))
train_db = train_db.shuffle(1000).map(pre_process).batch(64)
y_test = tf.squeeze(y_test, axis=1)
test_db = tf.data.Dataset.from_tensor_slices((x_test, y_test))
test_db = test_db.map(pre_process).batch(64)
sample = next(iter(train_db))
print(sample[0].shape, sample[1].shape, tf.reduce_min(sample[0]), tf.reduce_max(sample[0]))
def main():
# 卷积层 [b, 32, 32, 3] => [b, 1, 1, 512]
conv_net = Sequential(conv_layers)
# 全连接层 [b, 512] => [b, 100]
fc_net = Sequential([
# 此处代表输出参数
layers.Dense(256, activation=tf.nn.relu),
layers.Dense(128, activation=tf.nn.relu),
layers.Dense(100, activation=tf.nn.relu),
])
conv_net.build(input_shape=[None, 32, 32, 3])
fc_net.build(input_shape=[None, 512])
optimizer = optimizers.Adam(lr=1e-4)
# [1, 2] + [3, 4] => [1, 2, 3, 4]
variables = conv_net.trainable_variables + fc_net.trainable_variables
for epoch in range(50):
for step, (x, y) in enumerate(train_db):
with tf.GradientTape() as tape:
# [b, 32, 32, 3] => [b, 1, 1, 512]
out = conv_net(x)
# flatten => [b, 512]
out = tf.reshape(out, [-1, 512])
# [b, 512] => [b, 100]
logits = fc_net(out)
# [b] => [b, 100]
y_onehot = tf.one_hot(y, depth=100)
# compute loss
loss = tf.losses.categorical_crossentropy(y_onehot, logits, from_logits=True)
loss = tf.reduce_mean(loss)
grads = tape.gradient(loss, variables)
optimizer.apply_gradients(zip(grads, variables))
if step % 10 == 0:
print(epoch, step, "loss: ", float(loss))
total_num = 0
total_correct = 0
for x, y in test_db:
out = conv_net(x)
out = tf.reshape(out, [-1, 512])
logits = fc_net(out)
prob = tf.nn.softmax(logits, axis=1)
pred = tf.argmax(prob, axis=1)
pred = tf.cast(pred, dtype=tf.int32)
correct = tf.cast(tf.equal(pred, y), tf.float32)
correct = tf.reduce_sum(correct)
total_num += x.shape[0]
total_correct += correct
acc = total_correct / total_num
print("acc: ", acc)
if __name__ == '__main__':
main()
3. 经典卷积神经网络
- ImageNet
- LeNet-5
- AlexNet
- VGG
- GoogleNet