MCPcopy Create free account

hub / github.com/ali-vilab/dreamtalk / functions

Functions134 in github.com/ali-vilab/dreamtalk

⨍ Functions 134◇ Types & classes 40
↓ 15 callersMethod__init__
(self, input_nc, norm_layer=nn.BatchNorm2d, nonlinearity=nn.LeakyReLU(), use_spect=False)
generators/base_function.py:313
↓ 14 callersFunctionspectral_norm
use spectral normal layer to stable the training process
generators/base_function.py:151
↓ 5 callersMethod__init__
(self, d_model=512, nhead=8, num_encoder_layers=6, num_decoder_layers=6, dim_feedforward=204
core/networks/transformer.py:47
↓ 5 callersFunction_reset_parameters
(model)
core/utils.py:18
↓ 4 callersFunctionget_cfg_defaults
Get a yacs CfgNode object with default values for my_project.
configs/default.py:89
↓ 3 callersMethod__init__
(self, coeff_nc, descriptor_nc, layer)
generators/face_model.py:40
↓ 3 callersFunction_get_activation_fn
Return an activation function given a string
core/networks/transformer.py:32
↓ 3 callersFunction_get_clones
(module, N)
core/networks/transformer.py:42
↓ 3 callersFunctionget_network
(name: str)
core/networks/__init__.py:9
↓ 2 callersMethod__init__
( self, d_model=512, nhead=8, num_decoder_layers=3, dim_feedforward=20
core/networks/generator.py:216
↓ 2 callersMethod_latent_to_face3d
(self, latent)
core/networks/diffusion_net.py:75
↓ 2 callersFunctionget_decoder_network
( network_type, d_model, nhead, dim_feedforward, dropout, activation, normalize_be
core/networks/disentangle_decoder.py:13
↓ 1 callersMethod__init__
(self, cfg)
core/networks/diffusion_util.py:56
↓ 1 callersMethod__init__
(self, cond_planes, ratio, K, temperature=30, init_weight=True)
core/networks/dynamic_conv.py:9
↓ 1 callersMethod__init__
(self, decoder_layer, num_layers, norm=None, return_intermediate=False)
core/networks/dynamic_fc_decoder.py:119
↓ 1 callersMethod_get_sinusoid_encoding_table
Sinusoid position encoding table
core/networks/transformer.py:16
↓ 1 callersMethod_initialize_weights
(self)
core/networks/dynamic_conv.py:28
↓ 1 callersMethod_initialize_weights
(self)
core/networks/dynamic_conv.py:98
↓ 1 callersMethod_reset_parameters
(self)
core/networks/transformer.py:69
↓ 1 callersFunctioncompute_aspect_preserved_bbox
(bbox, increase_area, h, w)
core/utils.py:399
↓ 1 callersFunctioncrop_src_image
(src_img, save_img, increase_ratio, detector=None)
core/utils.py:433
↓ 1 callersMethodddim_sample
Args: audio (_type_): (B, L, W) or (B, L, W, C) style_clip (_type_): (B, L_clipmax, C_face3d) style_pad_
core/networks/diffusion_net.py:88
↓ 1 callersMethodddpm_sample
Args: audio (_type_): (B, L, W) or (B, L, W, C) style_clip (_type_): (B, L_clipmax, C_face3d) style_pad_
core/networks/diffusion_net.py:229
↓ 1 callersFunctionface3d_norm_to_raw
Args: face3d_norm (_type_): (B, L, C_face3d) exp_min (_type_): (C_face3d) exp_max (_type_): (C_face3d) Returns:
core/networks/diffusion_net.py:27
↓ 1 callersFunctionface3d_raw_to_norm
Args: face3d_raw (_type_): (B, L, C_face3d) exp_min (_type_): (C_face3d) exp_max (_type_): (C_face3d) Returns:
core/networks/diffusion_net.py:9
↓ 1 callersMethodforward_post
(self, src, src_mask = None, src_key_padding
core/networks/transformer.py:170
↓ 1 callersMethodforward_post
(self, tgt, memory, tgt_mask = None, memory_mask = None,
core/networks/transformer.py:233
↓ 1 callersMethodforward_post
( self, tgt, memory, style, tgt_mask=None, memory_mask=None,
core/networks/dynamic_fc_decoder.py:44
↓ 1 callersMethodforward_pre
(self, src, src_mask = None, src_key_padding_mask = None,
core/networks/transformer.py:185
↓ 1 callersMethodforward_pre
(self, tgt, memory, tgt_mask = None, memory_mask = None,
core/networks/transformer.py:256
↓ 1 callersFunctionget_diff_net
(cfg, device)
inference_for_demo_video.py:27
↓ 1 callersFunctionget_netG
(checkpoint_path, device)
generators/utils.py:20
↓ 1 callersFunctionget_pose_params
Get pose parameters from mat file Args: mat_path (str): path of mat file Returns: pose_params (numpy.ndarray): shape (L_vide
core/utils.py:288
↓ 1 callersMethodget_sigmas
(self, t, flexibility)
core/networks/diffusion_util.py:47
↓ 1 callersFunctionget_video_style
(video_name, style_type)
core/utils.py:24
↓ 1 callersFunctionget_video_style_clip
( video_name, video_root_dir, style_max_len, start_idx="random", dtype=torch.float32,
core/utils.py:92
↓ 1 callersFunctionget_wav2vec_audio_window
Args: audio_feat (np.ndarray): (N, 1024) start_idx (_type_): _description_ num_frames (_type_): _description_
core/utils.py:172
↓ 1 callersFunctioninference_one_video
( cfg, audio_path, style_clip_path, pose_path, output_path, diff_net, device,
inference_for_demo_video.py:57
↓ 1 callersFunctionmake_coordinate_grid
r"""obtain coordinate grid with the same size as the flow filed. Args: flow (tensor): Flow field obtained by the model Returns:
generators/flow_util.py:17
↓ 1 callersFunctionmish
Applies the mish function element-wise: mish(x) = x * tanh(softplus(x)) = x * tanh(ln(1 + exp(x))) See additional documentation for mish
core/networks/mish.py:12
↓ 1 callersFunctionobtain_seq_index
(index, num_frames, radius)
generators/utils.py:13
↓ 1 callersFunctionrender_video
exp: (N, 73)
generators/utils.py:39
↓ 1 callersMethodsample
( self, audio, style_clip, style_pad_mask, output_dim, flexibi
core/networks/diffusion_net.py:187
↓ 1 callersMethodshortcut
(self, x, z)
generators/base_function.py:146
↓ 1 callersFunctionsinusoidal_embedding
Args: timesteps (_type_): (B,) dim (_type_): (C_embed) Returns: _type_: (B, C_embed)
core/utils.py:311
Method__init__
( self, d_model=512, nhead=8, num_decoder_layers=3, dim_feedforward=20
core/networks/disentangle_decoder.py:78
Method__init__
(self, in_planes, out_planes, cond_planes, bias=True, K=4, temperature=30, ratio=4, init_weight=True)
core/networks/dynamic_linear.py:11
Method__init__
(self, num_steps, beta_1, beta_T, mode="linear")
core/networks/diffusion_util.py:10
Method__init__
(self, cfg, net, var_sched: VarianceSchedule)
core/networks/diffusion_net.py:46
Method__init__
( self, in_planes, out_planes, cond_planes, kernel_size, strid
core/networks/dynamic_conv.py:55
Method__init__
(self, input_dim)
core/networks/self_attention_pooling.py:13
Method__init__
(self, d_hid, n_position=200)
core/networks/transformer.py:10
Method__init__
(self, encoder_layer, num_layers, norm=None)
core/networks/transformer.py:91
Method__init__
(self, decoder_layer, num_layers, norm=None, return_intermediate=False)
core/networks/transformer.py:112
Method__init__
(self, d_model, nhead, dim_feedforward=2048, dropout=0.1, activation="relu", normalize_befor
core/networks/transformer.py:150
Method__init__
(self, d_model, nhead, dim_feedforward=2048, dropout=0.1, activation="relu", normalize_befor
core/networks/transformer.py:210
Method__init__
( self, d_model, nhead, d_style, dynamic_K, dynamic_ratio,
core/networks/dynamic_fc_decoder.py:9
Method__init__
( self, d_model=512, nhead=8, num_encoder_layers=6, dim_feedforward=20
core/networks/generator.py:75
Method__init__
( self, d_model=512, nhead=8, num_encoder_layers=6, dim_feedforward=20
core/networks/generator.py:133
Method__init__
Init method.
core/networks/mish.py:38
Method__init__
( self, mapping_net, warpping_net, editing_net, common )
generators/face_model.py:12
Method__init__
( self, image_nc, descriptor_nc, base_nc, max_nc, encoder
generators/face_model.py:66
Method__init__
( self, image_nc, descriptor_nc, layer, base_nc, max_nc,
generators/face_model.py:103
Method__init__
(self, n_out, affine=True)
generators/base_function.py:12
Method__init__
(self, image_nc, pose_nc, ngf, img_f, encoder_layers, decoder_layers, nonlinearity, use_spect)
generators/base_function.py:32
Method__init__
(self, image_nc, pose_nc, ngf, img_f, layers, nonlinearity=nn.LeakyReLU(), use_spect=False)
generators/base_function.py:44
Method__init__
(self, pose_nc, ngf, img_f, encoder_layers, decoder_layers, skip_connect=True, nonlinearity=
generators/base_function.py:66
Method__init__
(self, input_nc, output_nc, feature_nc, nonlinearity=nn.LeakyReLU(), use_spect=False)
generators/base_function.py:93
Method__init__
(self, input_nc, output_nc, hidden_nc, feature_nc, use_transpose=True, nonlinearity=nn.LeakyReLU(), use_spect=
generators/base_function.py:112
Method__init__
(self, norm_nc, feature_nc)
generators/base_function.py:160
Method__init__
(self, image_nc, ngf, img_f, layers, norm_layer=nn.BatchNorm2d, nonlinearity=nn.LeakyReLU(), use_spect=False)
generators/base_function.py:195
Method__init__
(self, image_nc, feature_nc, ngf, img_f, layers, num_block, norm_layer=nn.BatchNorm2d, nonlinearity=nn.LeakyRe
generators/base_function.py:217
Method__init__
(self, input_nc, output_nc, norm_layer=nn.BatchNorm2d, nonlinearity=nn.LeakyReLU(), use_spect=False)
generators/base_function.py:251
Method__init__
(self, input_nc, output_nc, norm_layer=nn.BatchNorm2d, nonlinearity=nn.LeakyReLU(), use_spect=False)
generators/base_function.py:267
Method__init__
(self, input_nc, output_nc, norm_layer=nn.BatchNorm2d, nonlinearity=nn.LeakyReLU(), use_spect=False)
generators/base_function.py:285
Method__init__
(self, num_block, input_nc, feature_nc, norm_layer=nn.BatchNorm2d, nonlinearity=nn.LeakyReLU(), use_spect=Fals
generators/base_function.py:299
Method__init__
(self, input_nc, feature_nc, norm_layer=nn.BatchNorm2d, nonlinearity=nn.LeakyReLU(), use_spect=False)
generators/base_function.py:331
Method__init__
(self, input_nc, output_nc, use_spect=False, tanh_or_sigmoid='tanh')
generators/base_function.py:354
Method_face3d_to_latent
(self, face3d)
core/networks/diffusion_net.py:63
Functionconvert_flow_to_deformation
r"""convert flow fields to deformations. Args: flow (tensor): Flow field obtained by the model Returns: deformation (tensor):
generators/flow_util.py:3
Functioncosine_loss
(a, v, y, logloss=nn.BCELoss())
core/utils.py:282
Methodforward
Args: content (_type_): (B, num_frames, window, C_dmodel) style_code (_type_): (B, C_dmodel) Returns:
core/networks/disentangle_decoder.py:169
Methodforward
Args: x (_type_): (L, B, C_in) cond (_type_): (B, C_style) Returns: _type_: (L, B, C_out)
core/networks/dynamic_linear.py:28
Methodforward
_summary_ Args: x_t (_type_): (B, L, C_face) t (_type_): (B,) dtype:float32 audio (_type_): (B, L, W)
core/networks/diffusion_util.py:89
Methodforward
Args: cond (_type_): (B, C_style) Returns: _type_: (B, K)
core/networks/dynamic_conv.py:38
Methodforward
Args: x (_type_): (B, C_in, L, 1) cond (_type_): (B, C_style) Returns: _type_: (B, C_out, L, 1)
core/networks/dynamic_conv.py:107
Methodforward
N: batch size, T: sequence length, H: Hidden dimension input: batch_rep : size (N, T, H) attention_weight:
core/networks/self_attention_pooling.py:18
Methodforward
(self, winsize)
core/networks/transformer.py:29
Methodforward
(self,opt, src, query_embed, pos_embed)
core/networks/transformer.py:74
Methodforward
(self, src, mask = None, src_key_padding_mask = None, pos = None)
core/networks/transformer.py:97
Methodforward
(self, tgt, memory, tgt_mask = None, memory_mask = None, tgt_key_padding_mask = None, memor
core/networks/transformer.py:119
Methodforward
(self, src, src_mask = None, src_key_padding_mask = None, p
core/networks/transformer.py:199
Methodforward
(self, tgt, memory, tgt_mask = None, memory_mask = None, tg
core/networks/transformer.py:279
Methodforward
( self, tgt, memory, style, tgt_mask=None, memory_mask=None,
core/networks/dynamic_fc_decoder.py:96
Methodforward
( self, tgt, memory, tgt_mask=None, memory_mask=None, tgt_key_
core/networks/dynamic_fc_decoder.py:126
Methodforward
Args: x (_type_): (B, num_frames, window, C_wav2vec) Returns: content: (B, num_frames, window, C_dmodel)
core/networks/generator.py:103
Methodforward
Args: x (_type_): (B, num_frames(L), C_exp) pad_mask: (B, num_frames) Returns: style_code: (B,
core/networks/generator.py:168
Methodforward
Args: content (_type_): (B, num_frames, window, C_dmodel) style_code (_type_): (B, C_dmodel) Returns:
core/networks/generator.py:255
Methodforward
Forward pass of the function.
core/networks/mish.py:44
Methodforward
( self, input_image, driving_source, stage=None )
generators/face_model.py:24
next →1–100 of 134, ranked by callers