MCPcopy Create free account

hub / github.com/JDAI-CV/FaceX-Zoo / functions

Functions1,054 in github.com/JDAI-CV/FaceX-Zoo

⨍ Functions 1,054◇ Types & classes 307
↓ 44 callersMethodnormalize
(self, factor)
addition_module/face_lightning/KDF/loss/ft.py:17
↓ 22 callersFunctionmake_layer
(block, num_of_layer, inc=64, outc=64, groups=1)
addition_module/DSDG/networks/generator.py:185
↓ 16 callersMethodbackward
(ctx, grad_output)
backbone/ReXNets.py:46
↓ 16 callersMethodcrop_image_by_mat
(self, image, landmarks)
face_sdk/core/image_cropper/arcface_cropper/FaceRecImageCropper.py:111
↓ 16 callersFunctionget_block
(in_channel, depth, num_units, stride = 2)
addition_module/face_lightning/KDF/backbone/ResNets.py:83
↓ 15 callersMethodupdate
(self, val, n=1)
addition_module/DSDG/misc/util.py:72
↓ 13 callersFunctionconv_dw
(inp, oup, stride, leaky=0.1)
face_sdk/models/network_def/retinaface_def.py:36
↓ 12 callersFunctionget_block
(in_channel, depth, num_units, stride = 2)
backbone/ResNets.py:83
↓ 10 callersMethodget_backbone
(self)
backbone/backbone_def.py:39
↓ 10 callersMethodget_head
(self)
head/head_def.py:36
↓ 9 callersMethodinference_on_image
(self, image)
face_sdk/core/model_handler/BaseModelHandler.py:25
↓ 9 callersMethodload_model
Should be overridden by all subclasses. Different models may have different configuration information, such as mean, so each model imp
face_sdk/core/model_loader/BaseModelLoader.py:47
↓ 9 callersMethodupdate
(self, val, n=1)
utils/AverageMeter.py:20
↓ 8 callersMethod_make_layer
(self, block, planes, blocks, stride=1, dilation=1, norm_layer=None, dropblock_prob=0.0, i
backbone/resnest/resnet.py:236
↓ 8 callersMethodget_lr
(self, epoch)
addition_module/DMUE/config.py:73
↓ 7 callersFunctionget_same_padding_conv2d
Chooses static padding if you have specified an image size, and dynamic padding otherwise. Static padding is necessary for ONNX exporting of mo
backbone/EfficientNets.py:197
↓ 7 callersMethodreset
(self)
addition_module/DSDG/misc/util.py:66
↓ 6 callersMethod__init__
(self, inplace=False)
backbone/TF_NAS.py:40
↓ 6 callersMethod__init__
(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.)
backbone/Swin_Transformer.py:20
↓ 6 callersMethod__init__
(self, in_channel, out_channel)
face_sdk/models/network_def/retinaface_def.py:48
↓ 6 callersMethod__init__
(self, inplace=True)
face_sdk/models/network_def/mobilev3_pfld.py:26
↓ 6 callersFunctioncv_crop
(image, landmarks, center, scale, resolution=256, center_shift=0)
addition_module/DMUE/preprocess/face_alignmenet/utils/utils.py:119
↓ 6 callersMethodget_backbone
(self)
addition_module/face_lightning/KDF/backbone/backbone_def.py:28
↓ 6 callersFunctionrgb2gray
(img)
addition_module/DSDG/misc/util.py:10
↓ 6 callersMethodtest
(self, model)
test_protocol/lfw/lfw_evaluator.py:33
↓ 6 callersMethodupdate
(self, val, n=1)
addition_module/DSDG/DUM/utils.py:23
↓ 5 callersMethod__init__
(self, block_args, global_params, image_size=None)
backbone/EfficientNets.py:659
↓ 5 callersMethod__init__
(self, in_channels, out_channels, kernel_size=3, stride=1, padding=1, type=1)
backbone/LightCNN.py:17
↓ 5 callersMethod__init__
(self, hdim=256, attack_type=4)
addition_module/DSDG/networks/generator.py:78
↓ 5 callersFunction_make_divisible
This function is taken from the original tf repo. It ensures that all layers have a channel number that is divisible by 8 It can be seen
backbone/GhostNet.py:31
↓ 5 callersFunctionconv3x3
3x3 convolution with padding
addition_module/DMUE/preprocess/face_alignmenet/core/models.py:8
↓ 5 callersFunctionconv_bn
(inp, oup, stride = 1, leaky = 0)
face_sdk/models/network_def/retinaface_def.py:16
↓ 5 callersFunctionconv_bn
(inp, oup, kernel_size, stride, padding=1, conv_layer=nn.Conv2d, norm_layer=nn.BatchNorm2d, nlin_layer=nn.ReLU
face_sdk/models/network_def/mobilev3_pfld.py:9
↓ 5 callersMethodconvert
(self, model, output_path)
addition_module/model_convertor/convert_to_onnx/model_convert.py:21
↓ 5 callersMethodload_model
The custom method to load a model. Args: model_path(str): the path of the weight file. Returns:
test_protocol/utils/model_loader.py:31
↓ 5 callersMethodload_model
The custom method to load a model. Args: model_path(str): the path of the weight file. Returns:
addition_module/face_lightning/KDF/test_protocol/utils/model_loader.py:31
↓ 5 callersMethodreset
(self)
utils/AverageMeter.py:14
↓ 5 callersFunctionround_filters
Calculate and round number of filters based on width multiplier. Use width_coefficient, depth_divisor and min_depth of global_params. Args
backbone/EfficientNets.py:81
↓ 4 callersMethod__init__
(self, input_channels, output_channels, stride=1)
backbone/AttentionNets.py:22
↓ 4 callersMethod__init__
(self, width=1.0, drop_ratio=0.2, feat_dim=512, out_h=7, out_w=7)
backbone/GhostNet.py:179
↓ 4 callersMethod__init__
(self, c, num_block, groups, kernel=(3, 3), stride=(1, 1), padding=(1, 1))
backbone/MobileFaceNets.py:59
↓ 4 callersMethod__init__
(self)
training_mode/siamese-triplet_training/pair_selector.py:20
↓ 4 callersMethod__init__
(self, c, num_block, groups, kernel=(3, 3), stride=(1, 1), padding=(1, 1))
addition_module/face_lightning/KDF/backbone/MobileFaceNets.py:59
↓ 4 callersMethod__init__
(self, c, num_block, groups, kernel=(3, 3), stride=(1, 1), padding=(1, 1))
face_sdk/models/network_def/mobilefacenet_def.py:63
↓ 4 callersMethod_make_layer
(self, block, num_blocks, in_channels, out_channels)
backbone/LightCNN.py:99
↓ 4 callersMethod_make_layer
(self, block, num_blocks, in_channels, out_channels)
backbone/LightCNN.py:151
↓ 4 callersMethod_make_layer
(self, block, planes, blocks, stride=1)
addition_module/DMUE/models/resnet_ibn_a.py:80
↓ 4 callersMethod_make_layer
(self, block, planes, blocks, stride=1)
addition_module/DMUE/models/resnet_ibn_multibranch.py:88
↓ 4 callersMethod_make_layer
(self, block, planes, blocks, stride=1)
addition_module/DMUE/models/resnet_multibranch.py:104
↓ 4 callersMethod_make_layer
(self, block, planes, blocks, stride=1)
addition_module/DMUE/models/resnet.py:97
↓ 4 callersMethod_make_layer
(self, block, planes, blocks, stride=1)
addition_module/DMUE/pretrain/backbones/resnet_ibn_a.py:106
↓ 4 callersMethod_make_layer
(self, block, planes, blocks, stride=1)
addition_module/DMUE/pretrain/backbones/resnet.py:100
↓ 4 callersMethod_make_layer
(self, block, num_blocks, in_channels, out_channels)
addition_module/DSDG/networks/light_cnn.py:66
↓ 4 callersMethod_make_stage
(self, planes, num_blocks, stride)
backbone/RepVGG.py:184
↓ 4 callersFunctioncrop_rfw
(rfw_root, rfw_lms_file, target_folder)
test_protocol/lfw/face_cropper/crop_rfw_by_arcface.py:13
↓ 4 callersFunctioncrop_rfw
(rfw_root, rfw_lms_file, target_folder)
addition_module/face_lightning/KDF/test_protocol/lfw/face_cropper/crop_rfw_by_arcface.py:13
↓ 4 callersMethoddetect
Detect all faces in PIL image and return bounding boxes and optional facial landmarks. This method is used by the forward method and is also
addition_module/DMUE/preprocess/mtcnn/mtcnn.py:272
↓ 4 callersFunctionfig2data
@brief Convert a Matplotlib figure to a 4D numpy array with RGBA channels and return it @param fig a matplotlib figure @return a numpy 3D
addition_module/DMUE/preprocess/face_alignmenet/utils/utils.py:338
↓ 4 callersFunctionget_err_threhold
(fpr, tpr, threshold)
addition_module/DSDG/DUM/utils.py:124
↓ 4 callersFunctionget_preds_fromhm
(hm, center=None, scale=None, rot=None)
addition_module/DMUE/preprocess/face_alignmenet/utils/utils.py:259
↓ 4 callersFunctionimresample
(img, sz)
addition_module/DMUE/preprocess/mtcnn/detect_face.py:304
↓ 4 callersFunctionload_model
(model, pretrained)
addition_module/DSDG/misc/util.py:39
↓ 4 callersMethodmake_layer
(self, block, channels, num_blocks, stride)
addition_module/DSDG/DUM/models/ResNet_u.py:74
↓ 4 callersFunctionmake_model
(cfg)
addition_module/DMUE/models/make_model.py:234
↓ 4 callersFunctionreduce_tensor
(tensor)
addition_module/DMUE/train_ddp.py:253
↓ 4 callersMethodtest
(self, model)
addition_module/face_lightning/KDF/test_protocol/lfw/lfw_evaluator.py:33
↓ 4 callersFunctiontransform
Transform a image by cv2.
addition_module/HSST/train_dataset.py:14
↓ 3 callersFunctionConvBNSwish
(out, in_channels, channels, kernel=1, stride=1, pad=0, num_group=1)
backbone/ReXNets.py:77
↓ 3 callersMethod__init__
(self, inplanes, planes, stride=1, downsample=None)
backbone/HRNet.py:38
↓ 3 callersMethod__init__
(self, num_layers, drop_ratio, mode='ir', feat_dim=512, out_h=7, out_w=7)
backbone/ResNets.py:112
↓ 3 callersMethod__init__
(self, in_channels, channels, se_ratio=12)
backbone/ReXNets.py:85
↓ 3 callersMethod__init__
(self, margin)
training_mode/siamese-triplet_training/losses.py:30
↓ 3 callersMethod__init__
(self, backbone, feat_dim, num_class, is_am)
addition_module/DMUE/pretrain/loss_def.py:127
↓ 3 callersMethod__init__
(self, pretrained=True)
addition_module/DMUE/preprocess/mtcnn/mtcnn.py:16
↓ 3 callersMethod__init__
(self, with_r=False)
addition_module/DMUE/preprocess/face_alignmenet/core/coord_conv.py:102
↓ 3 callersMethod__init__
(self, num_modules=1, end_relu=False, gray_scale=False, num_landmarks=68)
addition_module/DMUE/preprocess/face_alignmenet/core/models.py:146
↓ 3 callersMethod__init__
(self, num_layers, drop_ratio, mode='ir', feat_dim=512, out_h=7, out_w=7)
addition_module/face_lightning/KDF/backbone/ResNets.py:118
↓ 3 callersMethod__init__
(self, in_channels, out_channels, kernel_size=3, stride=1, padding=1, type=1)
addition_module/DSDG/networks/light_cnn.py:7
↓ 3 callersMethod__init__
(self, basic_conv=Conv2d_cd, theta=0.7)
addition_module/DSDG/DUM/models/CDCNs_u.py:55
↓ 3 callersMethod__init__
(self, in_channels=3, out_channels=3)
addition_module/face_mask_adding/FMA-3D/model/prnet.py:93
↓ 3 callersMethod_fuse_bn_tensor
(self, branch)
backbone/RepVGG.py:109
↓ 3 callersMethod_make_stage
(self, layer_config, num_inchannels, multi_scale_output=True)
backbone/HRNet.py:419
↓ 3 callersMethod_make_transition_layer
( self, num_channels_pre_layer, num_channels_cur_layer)
backbone/HRNet.py:366
↓ 3 callersFunctioncalculate_output_image_size
Calculates the output image size when using Conv2dSamePadding with a stride. Necessary for static padding. Thanks to mannatsingh for pointing t
backbone/EfficientNets.py:172
↓ 3 callersFunctionchannel_shuffle
(x, groups)
backbone/TF_NAS.py:14
↓ 3 callersFunctionconv_bn1X1
(inp, oup, stride, leaky=0)
face_sdk/models/network_def/retinaface_def.py:29
↓ 3 callersFunctionconv_bn_no_relu
(inp, oup, stride)
face_sdk/models/network_def/retinaface_def.py:23
↓ 3 callersFunctiondraw_gaussian
(image, point, sigma)
addition_module/DMUE/preprocess/face_alignmenet/utils/utils.py:42
↓ 3 callersMethodevaluate
Method to evaluate the input model. Args: model(obj): the loaded model.
test_protocol/utils/online_val.py:55
↓ 3 callersMethodextract_offline
Extract and save features. Args: feats_root(str): the path to save features. model(object): initialized model.
test_protocol/utils/extractor/feature_extractor.py:47
↓ 3 callersMethodextract_offline
Extract and save features. Args: feats_root(str): the path to save features. model(object): initialized model.
addition_module/face_lightning/KDF/test_protocol/utils/extractor/feature_extractor.py:47
↓ 3 callersMethodforward
(self, data, label)
training_mode/conventional_training/train.py:45
↓ 3 callersMethodforward
(self, data, label)
addition_module/face_lightning/KDF/training_mode/cls_training/train.py:47
↓ 3 callersFunctionkl_loss
(mu, logvar, prior_mu=0)
addition_module/DSDG/misc/util.py:21
↓ 3 callersMethodload_param
(self, cfg)
addition_module/DMUE/models/make_model.py:172
↓ 3 callersFunctionreparameterize
(mu, logvar)
addition_module/DSDG/misc/util.py:15
↓ 3 callersFunctionsave_checkpoint
(model_path, model, epoch, iteration, name)
addition_module/DSDG/misc/util.py:51
↓ 2 callersFunctionConvBNAct
(out, in_channels, channels, kernel=1, stride=1, pad=0, num_group=1, active=True, relu6=False)
backbone/ReXNets.py:68
↓ 2 callersMethod__init__
(self, block, layers, radix=1, groups=1, bottleneck_width=64, num_classes=1000, dilated=False
backbone/resnest/resnet.py:154
↓ 2 callersMethod__init__
Init MegaFaceEvaluator by some initial files. Args: facescrub_json_list(str): the facescrub list provided by official.
test_protocol/megaface/megaface_evaluator.py:35
next →1–100 of 1,054, ranked by callers