shuffle_image, index0 = processing.shuffle(shuffle_image, axis=1)
shuffle_image, index0 = processing.shuffle(shuffle_image, axis=0)
shuffle_image, index0 = processing.shuffle(shuffle_image, axis=1)
plt.subplot(1, 3, 3) # show
plt.imshow(shuffle_image, cmap='gray')
fluency = score.fluency(shuffle_image)
plt.title('shuffle %d' % fluency)
plt.axis('off')
plt.draw()
#%%
plt.figure(2)
temp_image = shuffle_image
for i in range(10):
recover = model.ImageRecover(temp_image)
t_image = []
fluency = []
new_image3, index = recover.direct_greed()
t_image.append(new_image3[0][0])
fluency.append(score.fluency(t_image[-1]))
for j in range(40):
new_image3, index = recover.svd_greed(j + 1)
t_image.append(new_image3[0][0])
fluency.append(score.fluency(t_image[-1]))
pos = np.array(fluency).argmin()
print(pos)
temp_image = t_image[pos]
plt.subplot(2, 5, i + 1)
plt.title('greed %d' % (fluency[pos]))
plt.imshow(temp_image, cmap='Greys_r')
plt.imshow(gray_image, cmap='gray')
fluency = score.fluency(gray_image)
plt.axis('off')
plt.title('%s-Gray %d' % (image_name, fluency))
# %% Shuffle
shuffle_image, index0 = processing.shuffle(gray_image, axis=0)
index0 = np.array(index0) # change to ndarray
plt.subplot(1, 3, 3) # show
plt.imshow(shuffle_image, cmap='gray')
fluency = score.fluency(shuffle_image)
plt.title('shuffle %d' % fluency)
plt.axis('off')
plt.draw()
# %% SVD image sort
recover = model.ImageRecover(shuffle_image)
new_image1, index = recover.svd_imsort()
new_image1 = new_image1[0][0]
# show
plt.figure(2)
plt.imshow(new_image1, cmap='gray')
fluency = score.fluency(new_image1)
k_coeff, k_distance = score.Kendal(index0[index], range(len(index)))
plt.title('SVD_imsort %d %.2f' % (fluency, k_coeff))
plt.axis('off')
plt.draw()
# %% SVD greed
plt.figure(3)
for i in range(1, 21):
plt.subplot(4, 5, i)
num_index = [0, 1, 2, 3, 4] # save 0,1,2,3,4,5,6,7,8,9
axis = 2
data = data_train.data[:1000]
data = np.array(data).transpose(0, 3, 1, 2)
# networks
data_loader_train = torch.utils.data.DataLoader(dataset=data_train, batch_size=Args.batch_size, shuffle=True)
data_loader_test = torch.utils.data.DataLoader(dataset=data_test, batch_size=Args.batch_size, shuffle=True)
cnn0 = models.CNN_CIFAR10().cuda() if Args.cuda else models.CNN_CIFAR10()
cnn1 = models.CNN_CIFAR10().cuda() if Args.cuda else models.CNN_CIFAR10()
cnn2 = models.CNN_CIFAR10().cuda() if Args.cuda else models.CNN_CIFAR10()
#%% Shuffle and Recover
data_shuffled, index0 = processing.shuffle(data, axis=axis)
index0 = np.array(index0) # change to ndarray
best_index = [float('inf'), []]
# direct greed
recover = model.ImageRecover(data_shuffled)
data_recover, index = recover.direct_greed()
fluency = score.fluency(data_recover)
k_coeff, k_distance = score.Kendal(index0[index], range(len(index)))
if fluency < best_index[0]:
best_index[1] = index
best_index[0] = fluency
if Args.show:
plt.ion()
plt.figure(1)
plt.subplot(1, 3, 1) # show
datasets.demo_show(data[num_index], datasets_name)
plt.title('Original'), plt.axis('off')
plt.subplot(1, 3, 2) # show
datasets.demo_show(data_shuffled[num_index], datasets_name)
plt.title('Shuffled'), plt.axis('off')