使用预训练的卷积网络作为 GAN 鉴别器

Question

我从 TF2.0 文档中提取了一些代码来从自定义数据集生成图像。代码是here

由于文档使用 Keras，我想我可能会将鉴别器网络更改为预训练网络，例如 InceptionV3，并且只训练顶层。我找到了this 代码（在一组新类上微调 InceptionV3）。我似乎无法弄清楚如何用另一个替换一个。我知道我试图用功能 API 替换顺序模式。但我想它们在某种程度上是相互关联的。不过，我不是 Keras 的常客。

我的问题是：如何将顺序模式中的自定义 CNN 替换为功能 API 中预训练的 CNN 以用作鉴别器？

编辑：如果有人有使用 GANEstimator 的示例，我会很高兴，因为我更习惯于 TF。

使用生成器生成随机图像

def make_generator_model():
    model = tf.keras.Sequential()
    model.add(layers.Dense(7*7*256, use_bias=False, input_shape=(100,)))
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())

    model.add(layers.Reshape((7, 7, 256)))
    assert model.output_shape == (None, 7, 7, 256) # Note: None is the batch size

    model.add(layers.Conv2DTranspose(128, (5, 5), strides=(1, 1), padding='same', use_bias=False))
    assert model.output_shape == (None, 7, 7, 128)
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())

    model.add(layers.Conv2DTranspose(64, (5, 5), strides=(2, 2), padding='same', use_bias=False))
    assert model.output_shape == (None, 14, 14, 64)
    model.add(layers.BatchNormalization())
    model.add(layers.LeakyReLU())

    model.add(layers.Conv2DTranspose(3, (5, 5), strides=(2, 2), padding='same', use_bias=False, activation='tanh'))
    assert model.output_shape == (None, 28, 28, 3)

    return model

generator = make_generator_model()
noise = tf.random.normal([1, 100])
generated_image = generator(noise, training=False)

当前的鉴别器和辅助器（输出 tf.Tensor([[-0.0003378]], shape=(1, 1), dtype=float32))

def make_discriminator_model():
    model = tf.keras.Sequential()
    model.add(layers.Conv2D(64, (5, 5), strides=(2, 2), padding='same',
                                     input_shape=[28, 28, 1]))
    model.add(layers.LeakyReLU())
    model.add(layers.Dropout(0.3))

    model.add(layers.Conv2D(128, (5, 5), strides=(2, 2), padding='same'))
    model.add(layers.LeakyReLU())
    model.add(layers.Dropout(0.3))

    model.add(layers.Flatten())
    model.add(layers.Dense(1))

    return model

cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True)

def discriminator_loss(real_output, fake_output):
    real_loss = cross_entropy(tf.ones_like(real_output), real_output)
    fake_loss = cross_entropy(tf.zeros_like(fake_output), fake_output)
    total_loss = real_loss + fake_loss
    return total_loss

discriminator = make_discriminator_model()
decision = discriminator(generated_image)
print (decision)

所需的鉴别器

def make_discriminator_model():
    # create the base pre-trained model
    model = InceptionV3(weights='imagenet', include_top=False)

    # ADD TOP LAYERS

    # FREEZE ALL LAYERS EXCEPT TOP LAYERS

    return model

# COMPILE

def discriminator_loss(real_output, fake_output):
    real_loss = ??? # Real Loss
    fake_loss = ??? # Fake loss
    total_loss = real_loss + fake_loss
    return total_loss

noise = tf.random.normal([1, 100])
generated_image = generator(noise, training=False)

discriminator = make_discriminator_model()
decision = discriminator(generated_image)
print (decision)

所有导入

  from __future__ import absolute_import, division, print_function, unicode_literals

try:
  # %tensorflow_version only exists in Colab.
  %tensorflow_version 2.x
except Exception:
  pass
import tensorflow as tf
print('TF version: {}'.format(tf.__version__))

import glob
import imageio
import matplotlib.pyplot as plt
import numpy as np
import os
import PIL
from PIL import Image
from tensorflow.keras import layers
import time

from IPython import display
from tensorflow.keras.preprocessing import image
from tensorflow.keras.applications import vgg16
import os.path
from tensorflow.keras.applications.inception_v3 import InceptionV3
from tensorflow.keras.preprocessing import image
from tensorflow.keras.models import Model
from tensorflow.keras.layers import Dense, GlobalAveragePooling2D
from tensorflow.keras import backend as K

编辑： 这是我最终得到的鉴别器！感谢@pandrey

def make_discriminator_model():
    pre_trained = tf.keras.applications.InceptionV3(
        weights='imagenet', include_top=False, input_shape=IMG_SHAPE
    )
    pre_trained.trainable = False  # mark all weights as non-trainable
    model = tf.keras.Sequential([pre_trained])
    model.add(layers.GlobalAveragePooling2D())
    model.add(layers.Dense(1))   
    return model

Answer 1

这是一种实现部分预训练和冻结模型构建的方法：

# Load the pre-trained model and freeze it.
pre_trained = tf.keras.applications.InceptionV3(
    weights='imagenet', include_top=False
)
pre_trained.trainable = False  # mark all weights as non-trainable
# Define a Sequential model, adding trainable layers on top of the previous.
model = tf.keras.Sequential([pre_trained])
# ADD TOP LAYERS HERE, just as you would in your example code
# You can check that the proper parameters are frozen or learnable with:
model.summary()

与您作为示例给出的损失函数相比，我认为您不需要更改损失函数。