-
Notifications
You must be signed in to change notification settings - Fork 8
/
performance.py
185 lines (134 loc) · 6.9 KB
/
performance.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
# Note: Doesn't work yet
#//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
#THIS SCRIPT HAS THE PURPOSE TO MAKE THE 'USUAL' DREAM (@script.py) FASTER (expiriment)
#//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
#SCROLL DOWN FOR SETTINGS (Google's Deep Dream Notebook code modified by me (@MartinKaptein) and converted to py script for convenience)
# imports and basic notebook setup
from cStringIO import StringIO
import numpy as np
import scipy.ndimage as nd
import PIL.Image
from IPython.display import clear_output, Image, display
from google.protobuf import text_format
import sys
#You will probably have to change this!
sys.path.append("/home/veli/src/caffe/distribute/python")
import caffe
# If your GPU supports CUDA and Caffe was built with CUDA support,
# uncomment the following to run Caffe operations on the GPU.
# caffe.set_mode_gpu()
# caffe.set_device(0) # select GPU device if multiple devices exist
def showarray(a, fmt='jpeg'):
a = np.uint8(np.clip(a, 0, 255))
f = StringIO()
PIL.Image.fromarray(a).save(f, fmt)
display(Image(data=f.getvalue()))
#Here you select the model
model_path = '/home/veli/src/caffe/models/bvlc_googlenet/' # substitute your path here >> this are my settings so you have to change them
net_fn = model_path + 'deploy.prototxt'
param_fn = model_path + 'bvlc_googlenet.caffemodel'
# Patching model to be able to compute gradients.
# Note that you can also manually add "force_backward: true" line to "deploy.prototxt".
model = caffe.io.caffe_pb2.NetParameter()
text_format.Merge(open(net_fn).read(), model)
model.force_backward = True
open('tmp.prototxt', 'w').write(str(model))
net = caffe.Classifier('tmp.prototxt', param_fn,
mean = np.float32([104.0, 116.0, 122.0]), # ImageNet mean, training set dependent
channel_swap = (2,1,0)) # the reference model has channels in BGR order instead of RGB
# a couple of utility functions for converting to and from Caffe's input image layout
def preprocess(net, img):
return np.float32(np.rollaxis(img, 2)[::-1]) - net.transformer.mean['data']
def deprocess(net, img):
return np.dstack((img + net.transformer.mean['data'])[::-1])
def objective_L2(dst):
dst.diff[:] = dst.data
#Jitter is best set to 0 in my experience
#SOME LITTLE SETTINGS
def make_step(net, step_size=1.6, end='inception_4c/output', jitter=0, clip=True, objective=objective_L2):
#function BAK def make_step(net, step_size=1.5, end='inception_4c/output', jitter=32, clip=True, objective=objective_L2):
'''Basic gradient ascent step.'''
src = net.blobs['data'] # input image is stored in Net's 'data' blob
dst = net.blobs[end]
ox, oy = np.random.randint(-jitter, jitter+1, 2)
src.data[0] = np.roll(np.roll(src.data[0], ox, -1), oy, -2) # apply jitter shift
net.forward(end=end)
objective(dst) # specify the optimization objective
net.backward(start=end)
g = src.diff[0]
# apply normalized ascent step to the input image
src.data[:] += step_size/np.abs(g).mean() * g
src.data[0] = np.roll(np.roll(src.data[0], -ox, -1), -oy, -2) # unshift image
if clip:
bias = net.transformer.mean['data']
src.data[:] = np.clip(src.data, -bias, 255-bias)
#///////////////////////////////////////////////////////////////////////////
#MAIN DEEPDREAM SETTINGS
#///////////////////////////////////////////////////////////////////////////
def deepdream(net, base_img, iter_n=11, octave_n=4, octave_scale=1.3, end='inception_4c/output', clip=True, **step_params):
#BACKUP high detail: def deepdream(net, base_img, iter_n=12, octave_n=6, octave_scale=1.6,end='inception_5b/pool_proj', clip=True, **step_params):
#deepdream(net, base_img, iter_n=10, octave_n=7, octave_scale=1.6,end='prob', clip=False, **step_params):
#function params>>net, base_img, iter_n=10, octave_n=4, octave_scale=1.4, end='inception_5b/5x5', clip=True, **step_params
# prepare base images for all octaves
octaves = [preprocess(net, base_img)]
for i in xrange(octave_n-1):
octaves.append(nd.zoom(octaves[-1], (1, 1.0/octave_scale,1.0/octave_scale), order=1))
src = net.blobs['data']
detail = np.zeros_like(octaves[-1]) # allocate image for network-produced details
for octave, octave_base in enumerate(octaves[::-1]):
h, w = octave_base.shape[-2:]
if octave > 0:
# upscale details from the previous octave
h1, w1 = detail.shape[-2:]
detail = nd.zoom(detail, (1, 1.0*h/h1,1.0*w/w1), order=1)
src.reshape(1,3,h,w) # resize the network's input image size
src.data[0] = octave_base+detail
for i in xrange(iter_n):
make_step(net, end=end, clip=clip, **step_params)
# visualization
vis = deprocess(net, src.data[0])
if not clip: # adjust image contrast if clipping is disabled
vis = vis*(255.0/np.percentile(vis, 99.98))
showarray(vis)
print octave, i, end, vis.shape
clear_output(wait=True)
# extract details produced on the current octave
detail = src.data[0]-octave_base
# returning the resulting image
return deprocess(net, src.data[0])
#SELECT HERE THE PICTURE YOU WANT TO DRAW THE DREAM ON:
#///////////////////////////////////////////////////////////////
img = np.float32(PIL.Image.open('source_pictures/tesla.jpg'))
#///////////////////////////////////////////////////////////////
#GUIDE IMAGE = ORIGINAL IMAGE WITH 1 DREAM (later 2) 'ON IT' (>> this is an experiment for total performance increase)
guide = deepdream(net, img)
#debug
print 'Guide is ready, now dreaming on (dreamt) guide'
#ADVANCED DREAM CONTROL AND FUNCTION DEF
#original>>end = 'inception_3b/output'
end = 'inception_4c/output'
h, w = guide.shape[:2]
src, dst = net.blobs['data'], net.blobs[end]
src.reshape(1,3,h,w)
src.data[0] = preprocess(net, guide)
net.forward(end=end)
guide_features = dst.data[0].copy()
def objective_guide(dst):
x = dst.data[0].copy()
y = guide_features
ch = x.shape[0]
x = x.reshape(ch,-1)
y = y.reshape(ch,-1)
A = x.T.dot(y) # compute the matrix of dot-products with guide features
dst.diff[0].reshape(ch,-1)[:] = y[:,A.argmax(1)] # select ones that match best
#deepdream(net, img, end=end, objective=objective_guide)
#GO
frame = img
counter = 0
for i in xrange(30): #change how many times you want to re-feed
frame = deepdream(net, frame, end=end, objective=objective_guide) #HERE HAPPENS THE MAJORITY OF THE WORK :)
PIL.Image.fromarray(np.uint8(frame)).save("output/%04d.jpg"%counter)
#Open the file we just created (again)
frame = np.float32(PIL.Image.open('output/%04d.jpg'%counter))
counter += 1
print 'Done!'