问题描述：

用keras实现DeepDream。我们将从一个在 ImageNet 上预训练的卷积神经网络开始。本次选择的模型是Inception模型

实现步骤：

下载数据集和设置DeepDream配置

from tensorflow import keras
import matplotlib.pyplot as pltbase_image_path = keras.utils.get_file("coast.jpg", origin="https://img-datasets.s3.amazonaws.com/coast.jpg")plt.axis("off")
plt.imshow(keras.utils.load_img(base_image_path))#instantiating a model
from tensorflow.keras.applications import inception_v3
model = inception_v3.InceptionV3(weights='imagenet',include_top=False)#配置各层对DeepDream损失的贡献
layer_settings = {"mixed4": 1.0,"mixed5": 1.5,"mixed6": 2.0,"mixed7": 2.5,
}
outputs_dict = dict([(layer.name, layer.output)for layer in [model.get_layer(name) for name in layer_settings.keys()]]
)
feature_extractor = keras.Model(inputs=model.inputs, outputs=outputs_dict)

定义最大化的损失和梯度上升过程

#定义损失函数
import tensorflow as tf
def compute_loss(input_image):features = feature_extractor(input_image)loss = tf.zeros(shape=())for name in features.keys():coeff = layer_settings[name]activation = features[name]loss += coeff * tf.reduce_mean(tf.square(activation[:, 2:-2, 2:-2, :]))return loss
#梯度上升过程
@tf.function
def gradient_ascent_step(image, learning_rate):with tf.GradientTape() as tape:tape.watch(image)loss = compute_loss(image)grads = tape.gradient(loss, image)grads = tf.math.l2_normalize(grads)image += learning_rate * gradsreturn loss, imagedef gradient_ascent_loop(image, iterations, learning_rate, max_loss=None):for i in range(iterations):loss, image = gradient_ascent_step(image, learning_rate)if max_loss is not None and loss > max_loss:breakprint(f"... Loss value at step {i}: {loss:.2f}")return image

在多个连续尺度上运行梯度上升

#在多个连续 上运行梯度上升
original_img = preprocess_image(base_image_path)
original_shape = original_img.shape[1:3]successive_shapes = [original_shape]
for i in range(1, num_octave):shape = tuple([int(dim / (octave_scale ** i)) for dim in original_shape])successive_shapes.append(shape)
successive_shapes = successive_shapes[::-1]shrunk_original_img = tf.image.resize(original_img, successive_shapes[0])img = tf.identity(original_img)
for i, shape in enumerate(successive_shapes):print(f"Processing octave {i} with shape {shape}")img = tf.image.resize(img, shape)img = gradient_ascent_loop(img, iterations=iterations, learning_rate=step, max_loss=max_loss)upscaled_shrunk_original_img = tf.image.resize(shrunk_original_img, shape)same_size_original = tf.image.resize(original_img, shape)lost_detail = same_size_original - upscaled_shrunk_original_imgimg += lost_detailshrunk_original_img = tf.image.resize(original_img, shape)keras.utils.save_img("DeepDream.png", deprocess_image(img.numpy()))

图像处理的辅助函数和超参数的设置

#hyperparameters
step = 20.
num_octave = 3
octave_scale = 1.4
iterations = 30
max_loss = 15.#图像处理方面
import numpy as npdef preprocess_image(image_path):img = keras.utils.load_img(image_path)img = keras.utils.img_to_array(img)img = np.expand_dims(img, axis=0)img = keras.applications.inception_v3.preprocess_input(img)return imgdef deprocess_image(img):img = img.reshape((img.shape[1], img.shape[2], 3))img /= 2.0img += 0.5img *= 255.img = np.clip(img, 0, 255).astype("uint8")return img

代码展示：

from tensorflow import keras
import matplotlib.pyplot as pltbase_image_path = keras.utils.get_file("coast.jpg", origin="https://img-datasets.s3.amazonaws.com/coast.jpg")plt.axis("off")
plt.imshow(keras.utils.load_img(base_image_path))#instantiating a model
from tensorflow.keras.applications import inception_v3
model = inception_v3.InceptionV3(weights='imagenet',include_top=False)#配置各层对DeepDream损失的贡献
layer_settings = {"mixed4": 1.0,"mixed5": 1.5,"mixed6": 2.0,"mixed7": 2.5,
}
outputs_dict = dict([(layer.name, layer.output)for layer in [model.get_layer(name) for name in layer_settings.keys()]]
)
feature_extractor = keras.Model(inputs=model.inputs, outputs=outputs_dict)#定义损失函数
import tensorflow as tf
def compute_loss(input_image):features = feature_extractor(input_image)loss = tf.zeros(shape=())for name in features.keys():coeff = layer_settings[name]activation = features[name]loss += coeff * tf.reduce_mean(tf.square(activation[:, 2:-2, 2:-2, :]))return loss#梯度上升过程
@tf.function
def gradient_ascent_step(image, learning_rate):with tf.GradientTape() as tape:tape.watch(image)loss = compute_loss(image)grads = tape.gradient(loss, image)grads = tf.math.l2_normalize(grads)image += learning_rate * gradsreturn loss, imagedef gradient_ascent_loop(image, iterations, learning_rate, max_loss=None):for i in range(iterations):loss, image = gradient_ascent_step(image, learning_rate)if max_loss is not None and loss > max_loss:breakprint(f"... Loss value at step {i}: {loss:.2f}")return image#hyperparameters
step = 20.
num_octave = 3
octave_scale = 1.4
iterations = 30
max_loss = 15.#图像处理方面
import numpy as npdef preprocess_image(image_path):img = keras.utils.load_img(image_path)img = keras.utils.img_to_array(img)img = np.expand_dims(img, axis=0)img = keras.applications.inception_v3.preprocess_input(img)return imgdef deprocess_image(img):img = img.reshape((img.shape[1], img.shape[2], 3))img /= 2.0img += 0.5img *= 255.img = np.clip(img, 0, 255).astype("uint8")return img#在多个连续 上运行梯度上升
original_img = preprocess_image(base_image_path)
original_shape = original_img.shape[1:3]successive_shapes = [original_shape]
for i in range(1, num_octave):shape = tuple([int(dim / (octave_scale ** i)) for dim in original_shape])successive_shapes.append(shape)
successive_shapes = successive_shapes[::-1]shrunk_original_img = tf.image.resize(original_img, successive_shapes[0])img = tf.identity(original_img)
for i, shape in enumerate(successive_shapes):print(f"Processing octave {i} with shape {shape}")img = tf.image.resize(img, shape)img = gradient_ascent_loop(img, iterations=iterations, learning_rate=step, max_loss=max_loss)upscaled_shrunk_original_img = tf.image.resize(shrunk_original_img, shape)same_size_original = tf.image.resize(original_img, shape)lost_detail = same_size_original - upscaled_shrunk_original_imgimg += lost_detailshrunk_original_img = tf.image.resize(original_img, shape)keras.utils.save_img("DeepDream.png", deprocess_image(img.numpy()))

运行截图：

参考：

《Python深度学习》

参考阅读：

tensorflow官方文档

https://colab.research.google.com/github/tensorflow/docs/blob/master/site/en/tutorials/generative/deepdream.ipynb

【深度学习】DeepDream的实现相关推荐

python深度学习--DeepDream
#DeepDream 是一种艺术性的图像修改技术,它用到了卷积神经网络学到的表示 ''' #DeepDream 算法与前面的卷积神经网络过滤器可视化技术几乎相同,都是反向运行一个cnn:对cnn的输入 ...
Python深度学习（DeepDream）--学习笔记（十九）
8.2 DeepDream DeepDream是一种艺术性的图像修改技术,它用到了卷积神经网络学到的表示.DeepDream由Google于2015年夏天首次发布,使用Caffe深度学习库编写实现.它 ...
Python深度学习之DeepDream
Deep Learning with Python 这篇文章是我学习<Deep Learning with Python>(第二版,François Chollet 著) 时写的系列笔记之 ...
Python深度学习（8）：DeepDream
算法简介 DeepDream是Google在2015年发出的一个比较好玩的算法.它的本意是去可视化神经网络.其思想简介就是,对于深度学习,以图片分类为例,我们最终的网络可以识别图像,是因为在训练过程中 ...
DeepDream:使用深度学习再造毕加索抽象风格艺术画
毕加索是近代最成功的艺术家,是抽象画派的开山师祖,而且凭借那些惊悚的抽象线条创造出来的画作非常挣钱.毕加索这种抽象创造能力能不能用计算机实现呢,随着深度学习的进一步发展,答案是肯定的. 抽象艺术能给人 ...
深度学习框架哪家强？MXNet称霸CNN、RNN和情感分析，TensorFlow仅擅长推断特征提取
深度学习框架哪家强:TensorFlow?Caffe?MXNet?Keras?PyTorch?对于这几大框架在运行各项深度任务时的性能差异如何,各位读者不免会有所好奇. 微软数据科学家Ilia Kar ...
【干货小铺】各种编程语言的深度学习库整理
点击上方"小白学视觉",选择加"星标"或"置顶" 重磅干货,第一时间送达本文转自|新机器视觉本文总结了Python.Matlab.CPP. ...
带你测试对比深度学习框架！TensorFlow,Keras,PyTorch...哪家强？(附数据集）
授权自AI科技大本营(ID: rgznai100) 本文长度为3556字,建议阅读7分钟亚马逊MXNet在CNN.RNN与NLP情感分析任务上性能强劲,而TensorFlow仅擅长于特征提取. 深度 ...
自学python要到什么程度-学好深度学习，Python 得达到什么程度？
如今,网络上的Python机器学习资源纷繁复杂,使得刚入门的小白们眼花缭乱.究竟从哪里开始?如何进行?云栖君给你推荐以下内容,相信读完你就会有自己的答案. "开始",是一个令人激动 ...

【深度学习】DeepDream的实现