"""VGG16 pretrained model and VGG Face model.
"""
"""
Copyright 2017 Parag K. Mital. See also NOTICE.md.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
"""
import tensorflow as tf
import json
import numpy as np
import matplotlib.pyplot as plt
from skimage.transform import resize as imresize
from .utils import download
[docs]def get_vgg_face_model():
"""Summary
Returns
-------
TYPE
Description
"""
download('https://s3.amazonaws.com/cadl/models/vgg_face.tfmodel')
with open("vgg_face.tfmodel", mode='rb') as f:
graph_def = tf.GraphDef()
try:
graph_def.ParseFromString(f.read())
except:
print('try adding PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python ' +
'to environment. e.g.:\n' +
'PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python ipython\n' +
'See here for info: ' +
'https://github.com/tensorflow/tensorflow/issues/582')
download('https://s3.amazonaws.com/cadl/models/vgg_face.json')
labels = json.load(open('vgg_face.json'))
return {
'graph_def': graph_def,
'labels': labels,
'preprocess': preprocess,
'deprocess': deprocess
}
[docs]def get_vgg_model():
"""Summary
Returns
-------
TYPE
Description
"""
download('https://s3.amazonaws.com/cadl/models/vgg16.tfmodel')
with open("vgg16.tfmodel", mode='rb') as f:
graph_def = tf.GraphDef()
try:
graph_def.ParseFromString(f.read())
except:
print('try adding PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python ' +
'to environment. e.g.:\n' +
'PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python ipython\n' +
'See here for info: ' +
'https://github.com/tensorflow/tensorflow/issues/582')
download('https://s3.amazonaws.com/cadl/models/synset.txt')
with open('synset.txt') as f:
labels = [(idx, l.strip()) for idx, l in enumerate(f.readlines())]
return {
'graph_def': graph_def,
'labels': labels,
'preprocess': preprocess,
'deprocess': deprocess
}
[docs]def preprocess(img, crop=True, resize=True, dsize=(224, 224)):
"""Summary
Parameters
----------
img : TYPE
Description
crop : bool, optional
Description
resize : bool, optional
Description
dsize : tuple, optional
Description
Returns
-------
TYPE
Description
"""
if img.dtype == np.uint8:
img = img / 255.0
if crop:
short_edge = min(img.shape[:2])
yy = int((img.shape[0] - short_edge) / 2)
xx = int((img.shape[1] - short_edge) / 2)
crop_img = img[yy: yy + short_edge, xx: xx + short_edge]
else:
crop_img = img
if resize:
norm_img = imresize(crop_img, dsize, preserve_range=True)
else:
norm_img = crop_img
return (norm_img).astype(np.float32)
[docs]def deprocess(img):
"""Summary
Parameters
----------
img : TYPE
Description
Returns
-------
TYPE
Description
"""
return np.clip(img * 255, 0, 255).astype(np.uint8)
# return ((img / np.max(np.abs(img))) * 127.5 +
# 127.5).astype(np.uint8)
[docs]def test_vgg():
"""Loads the VGG network and applies it to a test image.
"""
with tf.Session() as sess:
net = get_vgg_model()
tf.import_graph_def(net['graph_def'], name='vgg')
g = tf.get_default_graph()
names = [op.name for op in g.get_operations()]
input_name = names[0] + ':0'
x = g.get_tensor_by_name(input_name)
softmax = g.get_tensor_by_name(names[-2] + ':0')
og = plt.imread('bosch.png')
img = preprocess(og)[np.newaxis, ...]
res = np.squeeze(softmax.eval(feed_dict={
x: img,
'vgg/dropout_1/random_uniform:0': [[1.0]],
'vgg/dropout/random_uniform:0': [[1.0]]}))
print([(res[idx], net['labels'][idx])
for idx in res.argsort()[-5:][::-1]])
"""Let's visualize the network's gradient activation
when backpropagated to the original input image. This
is effectively telling us which pixels contribute to the
predicted class or given neuron"""
features = [name for name in names if 'BiasAdd' in name.split()[-1]]
from math import sqrt, ceil
n_plots = ceil(sqrt(len(features) + 1))
fig, axs = plt.subplots(n_plots, n_plots)
plot_i = 0
axs[0][0].imshow(img[0])
for feature_i, featurename in enumerate(features):
plot_i += 1
feature = g.get_tensor_by_name(featurename + ':0')
neuron = tf.reduce_max(feature, 1)
saliency = tf.gradients(tf.reduce_sum(neuron), x)
neuron_idx = tf.arg_max(feature, 1)
this_res = sess.run([saliency[0], neuron_idx], feed_dict={
x: img,
'vgg/dropout_1/random_uniform:0': [[1.0]],
'vgg/dropout/random_uniform:0': [[1.0]]})
grad = this_res[0][0] / np.max(np.abs(this_res[0]))
ax = axs[plot_i // n_plots][plot_i % n_plots]
ax.imshow((grad * 127.5 + 127.5).astype(np.uint8))
ax.set_title(featurename)
"""Deep Dreaming takes the backpropagated gradient activations
and simply adds it to the image, running the same process again
and again in a loop. There are many tricks one can add to this
idea, such as infinitely zooming into the image by cropping and
scaling, adding jitter by randomly moving the image around, or
adding constraints on the total activations."""
og = plt.imread('street.png')
crop = 2
img = preprocess(og)[np.newaxis, ...]
layer = g.get_tensor_by_name(features[3] + ':0')
n_els = layer.get_shape().as_list()[1]
neuron_i = np.random.randint(1000)
layer_vec = np.zeros((1, n_els))
layer_vec[0, neuron_i] = 1
neuron = tf.reduce_max(layer, 1)
saliency = tf.gradients(tf.reduce_sum(neuron), x)
for it_i in range(3):
print(it_i)
this_res = sess.run(saliency[0], feed_dict={
x: img,
layer: layer_vec,
'vgg/dropout_1/random_uniform:0': [[1.0]],
'vgg/dropout/random_uniform:0': [[1.0]]})
grad = this_res[0] / np.mean(np.abs(grad))
img = img[:, crop:-crop - 1, crop:-crop - 1, :]
img = imresize(img[0], (224, 224))[np.newaxis]
img += grad
plt.imshow(deprocess(img[0]))
[docs]def test_vgg_face():
"""Loads the VGG network and applies it to a test image.
"""
with tf.Session() as sess:
net = get_vgg_face_model()
x = tf.placeholder(tf.float32, [1, 224, 224, 3], name='x')
tf.import_graph_def(net['graph_def'], name='vgg',
input_map={'Placeholder:0': x})
g = tf.get_default_graph()
names = [op.name for op in g.get_operations()]
og = plt.imread('bricks.png')[..., :3]
img = preprocess(og)[np.newaxis, ...]
plt.imshow(img[0])
plt.show()
"""Let's visualize the network's gradient activation
when backpropagated to the original input image. This
is effectively telling us which pixels contribute to the
predicted class or given neuron"""
features = [name for name in names if 'BiasAdd' in name.split()[-1]]
from math import sqrt, ceil
n_plots = ceil(sqrt(len(features) + 1))
fig, axs = plt.subplots(n_plots, n_plots)
plot_i = 0
axs[0][0].imshow(img[0])
for feature_i, featurename in enumerate(features):
plot_i += 1
feature = g.get_tensor_by_name(featurename + ':0')
neuron = tf.reduce_max(feature, 1)
saliency = tf.gradients(tf.reduce_sum(neuron), x)
neuron_idx = tf.arg_max(feature, 1)
this_res = sess.run([saliency[0], neuron_idx], feed_dict={x: img})
grad = this_res[0][0] / np.max(np.abs(this_res[0]))
ax = axs[plot_i // n_plots][plot_i % n_plots]
ax.imshow((grad * 127.5 + 127.5).astype(np.uint8))
ax.set_title(featurename)
plt.waitforbuttonpress()
"""Deep Dreaming takes the backpropagated gradient activations
and simply adds it to the image, running the same process again
and again in a loop. There are many tricks one can add to this
idea, such as infinitely zooming into the image by cropping and
scaling, adding jitter by randomly moving the image around, or
adding constraints on the total activations."""
og = plt.imread('street.png')
crop = 2
img = preprocess(og)[np.newaxis, ...]
layer = g.get_tensor_by_name(features[3] + ':0')
n_els = layer.get_shape().as_list()[1]
neuron_i = np.random.randint(1000)
layer_vec = np.zeros((1, n_els))
layer_vec[0, neuron_i] = 1
neuron = tf.reduce_max(layer, 1)
saliency = tf.gradients(tf.reduce_sum(neuron), x)
for it_i in range(3):
print(it_i)
this_res = sess.run(saliency[0], feed_dict={
x: img,
layer: layer_vec,
'vgg/dropout_1/random_uniform:0': [[1.0]],
'vgg/dropout/random_uniform:0': [[1.0]]})
grad = this_res[0] / np.mean(np.abs(grad))
img = img[:, crop:-crop - 1, crop:-crop - 1, :]
img = imresize(img[0], (224, 224))[np.newaxis]
img += grad
plt.imshow(deprocess(img[0]))
if __name__ == '__main__':
test_vgg_face()