背景
我一直在玩Deep Dream和Inceptionism，使用Caffe框架来可视化GoogLeNet的层，这是为Imagenet项目构建的架构，是一个用于视觉对象识别的大型可视化数据库。
你可以在这里找到Imagenet: Imagenet 1000类。
为了探索建筑并产生“梦想”，我使用了三本笔记本:
https://github.com/google/deepdream/blob/master/dream.ipynb https://github.com/kylemcdonald/deepdream/blob/master/dream.ipynb https://github.com/auduno/deepdraw/blob/master/deepdraw.ipynb
这里的基本思想是从模型或“指南”图像中指定层的每个通道中提取一些特征。
然后，我们输入一张我们想要修改的图像到模型中，并提取指定的同一层(每个八度)中的特征， 增强最佳匹配特征，即两个特征向量的最大点积。
到目前为止，我已经设法修改输入图像和控制梦境使用以下方法:
(a)应用图层作为输入图像优化的“end”目标。(参见特征可视化) (b)使用第二图像对输入图像指导优化目标。 (c)可视化由噪声生成的Googlenet模型类。
然而，我想要达到的效果介于这些技术之间，我还没有找到任何文档、论文或代码。
期望结果(不是待回答问题的一部分)
让一个属于给定“end”层的单一类或单元(a)引导优化目标(b)，并在输入图像上可视化该类(c):
一个class = 'face'和input_image = 'clouds.jpg'的例子:
请注意:上图是使用人脸识别模型生成的，该模型没有在Imagenet数据集上进行训练。仅供演示之用。
工作代码
方法(一)
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 matplotlib as plt
import caffe
model_name = 'GoogLeNet'
model_path = 'models/dream/bvlc_googlenet/' # substitute your path here
net_fn = model_path + 'deploy.prototxt'
param_fn = model_path + 'bvlc_googlenet.caffemodel'
model = caffe.io.caffe_pb2.NetParameter()
text_format.Merge(open(net_fn).read(), model)
model.force_backward = True
open('models/dream/bvlc_googlenet/tmp.prototxt', 'w').write(str(model))
net = caffe.Classifier('models/dream/bvlc_googlenet/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
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()))
# 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
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)
def deepdream(net, base_img, iter_n=20, octave_n=4, octave_scale=1.4,
end='inception_4c/output', 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])
我运行上面的代码:
end = 'inception_4c/output'
img = np.float32(PIL.Image.open('clouds.jpg'))
_=deepdream(net, img)
方法(b)
"""
Use one single image to guide
the optimization process.
This affects the style of generated images
without using a different training set.
"""
def dream_control_by_image(optimization_objective, end):
# this image will shape input img
guide = np.float32(PIL.Image.open(optimization_objective))
showarray(guide)
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)
然后运行上面的代码:
end = 'inception_4c/output'
# image to be modified
img = np.float32(PIL.Image.open('img/clouds.jpg'))
guide_image = 'img/guide.jpg'
dream_control_by_image(guide_image, end)
问题
现在，我尝试访问单个类的失败方法，对类的矩阵进行热编码，并专注于一个(到目前为止还没有效果):
def objective_class(dst, class=50):
# according to imagenet classes
#50: 'American alligator, Alligator mississipiensis',
one_hot = np.zeros_like(dst.data)
one_hot.flat[class] = 1.
dst.diff[:] = one_hot.flat[class]
明确一点:这个问题不是关于梦想的代码，这是一个有趣的背景和已经工作的代码，但它只是关于最后一段的问题:有人能指导我如何从ImageNet获得一个所选类的图像(以类#50:“美洲短短鳄，密西西比短短鳄”为例)(这样我就可以将它们作为输入-与云图一起-创建一个梦想的图像)?