MCPcopy Create free account

hub / github.com/SystemErrorWang/White-box-Cartoonization / functions

Functions72 in github.com/SystemErrorWang/White-box-Cartoonization

⨍ Functions 72◇ Types & classes 2
↓ 12 callersMethodconv_layer
(self, bottom, name)
train_code/loss.py:70
↓ 12 callersFunctiontf_box_filter
(x, r)
test_code/guided_filter.py:7
↓ 7 callersFunctiontf_box_filter
(x, r)
train_code/guided_filter.py:11
↓ 5 callersMethodbuild_conv4_4
(self, rgb, include_fc=False)
train_code/loss.py:24
↓ 4 callersMethodmax_pool
(self, bottom, name)
train_code/loss.py:66
↓ 3 callersMethodbuild_region_pairs
(self)
train_code/selective_search/structure.py:36
↓ 3 callersMethodbuild_regions
(self)
train_code/selective_search/structure.py:15
↓ 3 callersMethodcalculate_similarity_for_new_region
(self)
train_code/selective_search/structure.py:115
↓ 3 callersMethodget_highest_similarity
(self)
train_code/selective_search/structure.py:66
↓ 3 callersFunctionguided_filter
(x, y, r, eps=1e-2)
train_code/guided_filter.py:21
↓ 3 callersMethodmerge_region
(self, i, j)
train_code/selective_search/structure.py:69
↓ 3 callersMethodremove_similarities
(self, i, j)
train_code/selective_search/structure.py:100
↓ 3 callersFunctionswitch_color_space
RGB to target color space conversion. I: the intensity (grey scale), Lab, rgI: the rg channels of normalized RGB plus intensi
train_code/selective_search/util.py:22
↓ 2 callersMethod_find_neighbors
Parameters ---------- label : int label of the region Returns ------- nei
train_code/selective_search/structure.py:48
↓ 2 callersMethodget_bias
(self, name)
train_code/loss.py:104
↓ 2 callersFunctionlabel2rgb
(label_field, image, kind='mix', bg_label=-1, bg_color=(0, 0, 0))
train_code/utils.py:40
↓ 2 callersFunctionlabel2rgb
(label_field, image, kind='avg', bg_label=-1, bg_color=(0, 0, 0))
train_code/selective_search/adaptive_color.py:4
↓ 2 callersMethodnum_regions
(self)
train_code/selective_search/structure.py:130
↓ 2 callersFunctionresblock
(inputs, out_channel=32, name='resblock')
train_code/network.py:16
↓ 1 callersFunction_calculate_color_sim
Calculate color similarity using histogram intersection
train_code/selective_search/measure.py:4
↓ 1 callersFunction_calculate_fill_sim
Fill similarity measures how well ri and rj fit into each other. BBij is the bounding box around ri and rj. fill(ri, rj) = 1
train_code/selective_search/measure.py:28
↓ 1 callersFunction_calculate_size_sim
Size similarity boosts joint between small regions, which prevents a single region from engulfing other blobs one by one. si
train_code/selective_search/measure.py:18
↓ 1 callersFunction_calculate_texture_sim
Calculate texture similarity using histogram intersection
train_code/selective_search/measure.py:11
↓ 1 callersFunctionarg_parser
()
train_code/train.py:24
↓ 1 callersFunctionarg_parser
()
train_code/pretrain.py:23
↓ 1 callersFunctioncartoonize
(load_folder, save_folder, model_path)
test_code/cartoonize.py:25
↓ 1 callersMethodget_conv_filter
(self, name)
train_code/loss.py:101
↓ 1 callersMethodget_fc_weight
(self, name)
train_code/loss.py:107
↓ 1 callersFunctionguided_filter
(x, y, r, eps=1e-2)
test_code/guided_filter.py:18
↓ 1 callersMethodis_empty
(self)
train_code/selective_search/structure.py:126
↓ 1 callersFunctionload_strategy
(mode)
train_code/selective_search/util.py:53
↓ 1 callersFunctionoversegmentation
Generating various starting regions using the method of Felzenszwalb. k effectively sets a scale of observation, in that
train_code/selective_search/util.py:8
↓ 1 callersFunctionresblock
(inputs, out_channel=32, name='resblock')
test_code/network.py:7
↓ 1 callersFunctionresize_crop
(image)
test_code/cartoonize.py:11
↓ 1 callersFunctionspectral_norm
(w, iteration=1)
train_code/layers.py:23
↓ 1 callersFunctiontrain
(args)
train_code/train.py:40
↓ 1 callersFunctiontrain
(args)
train_code/pretrain.py:38
Method__init__
(self, vgg19_npy_path=None)
train_code/loss.py:18
Method__init__
(self, img, img_seg, sim_strategy)
train_code/selective_search/structure.py:9
Functionadaptive_instance_norm
(content, style, epsilon=1e-5)
train_code/layers.py:13
Functionbox_filter
(boxes, min_size=20, max_ratio=None, topN=None)
train_code/selective_search/core.py:93
Functioncalculate_color_hist
Calculate colour histogram for the region. The output will be an array with n_BINS * n_color_channels. The number of channel
train_code/selective_search/measure.py:41
Functioncalculate_sim
Calculate similarity between region ri and rj using diverse combinations of similarity measures. C: color, T: texture, S: siz
train_code/selective_search/measure.py:101
Functioncalculate_texture_hist
Use LBP for now, enlightened by AlpacaDB's implementation. Plan to switch to Gaussian derivatives as the paper in future vers
train_code/selective_search/measure.py:80
Functioncolor_shift
(image1, image2, mode='uniform')
train_code/utils.py:22
Functioncolor_ss_map
(image, seg_num=200, power=1, color_space='Lab', k=10, sim_strategy='CTSF')
train_code/utils.py:73
Functioncolor_ss_map
(image, color_space='Lab', k=10, sim_strategy='CTSF', seg_num=200, power=1)
train_code/selective_search/batch_ss.py:15
Functionconv_spectral_norm
(x, channel, k_size, stride=1, name='conv_snorm')
train_code/layers.py:55
Functiondisc_bn
(x, scale=1, channel=32, is_training=True, name='discriminator', patch=True, reuse=False)
train_code/network.py:101
Functiondisc_ln
(x, channel=32, is_training=True, name='discriminator', patch=True, reuse=False)
train_code/network.py:149
Functiondisc_sn
(x, scale=1, channel=32, patch=True, name='discriminator', reuse=False)
train_code/network.py:126
Functionfast_guided_filter
(lr_x, lr_y, hr_x, r=1, eps=1e-8)
test_code/guided_filter.py:42
Methodfc_layer
(self, bottom, name)
train_code/loss.py:84
Functiongan_loss
(discriminator, real, fake, scale=1,channel=32, patch=False, name='discriminator')
train_code/loss.py:149
Functiongenerate_lbp_image
(img)
train_code/selective_search/measure.py:66
Functiongenerator
(inputs, channel=32, num_blocks=4, name='generator', reuse=False)
train_code/network.py:30
Functionload_image_list
(data_dir)
train_code/utils.py:124
Functionlsgan_loss
(discriminator, real, fake, scale=1, channel=32, patch=False, name='discriminator')
train_code/loss.py:164
Functionnext_batch
(filename_list, batch_size)
train_code/utils.py:132
Functionprocess_slic
(image)
train_code/utils.py:111
Functionselective_adacolor
(batch_image, seg_num=200, power=1)
train_code/utils.py:101
Functionselective_adacolor
(batch_image, seg_num=200, power=1)
train_code/selective_search/batch_ss.py:43
Functionselective_search
Selective Search in Python
train_code/selective_search/core.py:65
Functionselective_search_one
Selective Search using single diversification strategy Parameters ---------- im_orig : ndarray Original image
train_code/selective_search/core.py:9
Functionself_attention
(inputs, name='attention', reuse=False)
train_code/layers.py:66
Functionsimple_superpixel
(batch_image, seg_num=200)
train_code/utils.py:109
Functiontotal_variation_loss
(image, k_size=1)
train_code/loss.py:177
Functionunet_generator
(inputs, channel=32, num_blocks=4, name='generator', reuse=False)
train_code/network.py:62
Functionunet_generator
(inputs, channel=32, num_blocks=4, name='generator', reuse=False)
test_code/network.py:22
Functionvggloss_4_4
(image_a, image_b)
train_code/loss.py:112
Functionwgan_loss
(discriminator, real, fake, patch=True, channel=32, name='discriminator', lambda_=2)
train_code/loss.py:124
Functionwrite_batch_image
(image, save_dir, name, n)
train_code/utils.py:147