Functions — github.com/KlingAIResearch/LivePortrait

hub / github.com/KlingAIResearch/LivePortrait / functions

Functions617 in github.com/KlingAIResearch/LivePortrait

⨍ Functions 617◇ Types & classes 112

↓ 81 callersMethodto

(self, device)

src/utils/dependencies/XPose/util/misc.py:404

↓ 69 callersMethodget

(self, img_bgr, **kwargs)

src/utils/face_analysis_diy.py:41

↓ 49 callersFunctionprint

(*args, **kwargs)

src/utils/dependencies/XPose/util/misc.py:519

↓ 29 callersMethodmax

(self)

src/utils/dependencies/XPose/util/misc.py:86

↓ 25 callersFunctionbasename

a/b/c.jpg -> c

src/utils/helper.py:60

↓ 23 callersMethodrun

(self, input_image, instance_text_prompt, keypoint_text_example, box_threshold, IoU_threshold)

src/utils/animal_landmark_runner.py:106

↓ 21 callersMethodupdate

(self, *args, **kwargs)

src/utils/dependencies/XPose/util/addict.py:103

↓ 20 callersFunctionmake_abs_path

(fn)

src/config/base_config.py:12

↓ 15 callersFunctionload_model

(ckpt_path, model_config, device, model_type)

src/utils/helper.py:131

↓ 13 callersMethod__init__

(self, dim)

src/modules/util.py:333

↓ 13 callersMethodget

(self, img, face)

src/utils/dependencies/insightface/model_zoo/landmark.py:80

↓ 12 callersMethodcopy

(self)

src/utils/dependencies/XPose/util/addict.py:90

↓ 12 callersFunctioninverse_sigmoid

(x, eps=1e-3)

src/utils/dependencies/XPose/util/misc.py:689

↓ 10 callersMethoddeepcopy

(self)

src/utils/dependencies/XPose/util/addict.py:93

↓ 9 callersFunctionget_rotation_matrix

the input is in degree

src/utils/camera.py:31

↓ 9 callersFunctionheadpose_pred_to_degree

pred: (bs, 66) or (bs, 1) or others

src/utils/camera.py:14

↓ 9 callersFunctionimages2video

(images, wfp, **kwargs)

src/utils/video.py:25

↓ 8 callersFunctionload_description

(fp)

src/utils/helper.py:175

↓ 7 callersFunction_get_activation_fn

Return an activation function given a string

src/utils/dependencies/XPose/models/UniPose/utils.py:177

↓ 7 callersMethodextract_feature_3d

get the appearance feature of the image by F x: Bx3xHxW, normalized to 0~1

src/live_portrait_wrapper.py:122

↓ 7 callersMethodget_kp_info

get the implicit keypoint information x: Bx3xHxW, normalized to 0~1 flag_refine_info: whether to trandform the pose to degrees and th

src/live_portrait_wrapper.py:131

↓ 7 callersMethodstitching

conduct the stitching kp_source: Bxnum_kpx3 kp_driving: Bxnum_kpx3

src/live_portrait_wrapper.py:253

↓ 7 callersMethodupdate_config

(self, user_args)

src/utils/cropper.py:88

↓ 6 callersMethod__init__

(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.)

src/utils/dependencies/XPose/models/UniPose/swin_transformer.py:17

↓ 6 callersFunctionadd_audio_to_video

(silent_video_path: str, audio_video_path: str, output_video_path: str)

src/utils/video.py:189

↓ 6 callersFunctionconcat_feat

kp_source: (bs, k, 3) kp_driving: (bs, k, 3) Return: (bs, 2k*3)

src/utils/helper.py:110

↓ 6 callersFunctionconcat_frames

(driving_image_lst, source_image_lst, I_p_lst)

src/utils/video.py:84

↓ 6 callersFunctioncontiguous

(obj)

src/utils/io.py:32

↓ 6 callersFunctionexec_cmd

(cmd)

src/utils/video.py:21

↓ 6 callersMethodget

(self, key)

src/utils/dependencies/XPose/models/registry.py:34

↓ 6 callersFunctionpaste_back

paste back the image

src/utils/crop.py:417

↓ 6 callersFunctionprepare_paste_back

prepare mask for later image paste back

src/utils/crop.py:410

↓ 6 callersFunctionresize_to_limit

ajust the size of the image so that the maximum dimension does not exceed max_dim, and the width and the height of the image are multiples of n.

src/utils/io.py:38

↓ 5 callersMethodcalc_combined_lip_ratio

(self, c_d_lip_i, source_lmk)

src/live_portrait_wrapper.py:328

↓ 5 callersFunctiondct2device

(dct: dict, device)

src/utils/helper.py:101

↓ 5 callersFunctionget_fps

(filepath, default_fps=25)

src/utils/video.py:138

↓ 5 callersFunctionload_video

(video_info, n_frames=-1)

src/utils/io.py:19

↓ 5 callersFunctionpad

(image, target, padding)

src/utils/dependencies/XPose/transforms.py:213

↓ 5 callersMethodparse_output

construct the output as standard return: 1xHxWx3, uint8

src/live_portrait_wrapper.py:300

↓ 5 callersMethodprepare_videos

construct the input as standard imgs: NxBxHxWx3, uint8

src/live_portrait_wrapper.py:104

↓ 5 callersMethodretarget_lip

kp_source: BxNx3 lip_close_ratio: Bx2 Return: Bx(3*num_kp)

src/live_portrait_wrapper.py:227

↓ 5 callersFunctionsmooth

(x_d_lst, shape, device, observation_variance=3e-7, process_variance=1e-5)

src/utils/filter.py:8

↓ 5 callersFunctionupdate_args

update the args according to user inputs

src/gradio_pipeline.py:28

↓ 5 callersMethodwarp_decode

get the image after the warping of the implicit keypoints feature_3d: Bx32x16x64x64, feature volume kp_source: BxNx3 kp_drivi

src/live_portrait_wrapper.py:276

↓ 5 callersMethodwrite

(self, image)

src/utils/video.py:122

↓ 4 callersMethod__init__

(self, d_model=256, nhead=8, num_encoder_layers=6, num_decoder_layers=6, dim_feedforward=1024

src/utils/dependencies/XPose/models/UniPose/transformer_deformable.py:26

↓ 4 callersMethod__init__

_summary_ Args: encoder_layer (_type_): _description_ num_layers (_type_): _description_ norm (_type_, op

src/utils/dependencies/XPose/models/UniPose/deformable_transformer.py:527

↓ 4 callersMethod__setattr__

(self, name, value)

src/utils/dependencies/XPose/util/config.py:332

↓ 4 callersMethod__setattr__

(self, name, value)

src/utils/dependencies/XPose/util/addict.py:25

↓ 4 callersFunction_get_clones

(module, N, layer_share=False)

src/utils/dependencies/XPose/models/UniPose/deformable_transformer.py:1135

↓ 4 callersMethod_hook

(cls, item)

src/utils/dependencies/XPose/util/addict.py:59

↓ 4 callersMethod_shape

(self, tensor: torch.Tensor, seq_len: int, bsz: int)

src/utils/dependencies/XPose/models/UniPose/fuse_modules.py:127

↓ 4 callersFunction_transform_img

conduct similarity or affine transformation to the image, do not do border operation! img: M: 2x3 matrix or 3x3 matrix dsize: target shap

src/utils/crop.py:19

↓ 4 callersMethodcalc_lmks_from_cropped_video

Tracking based landmarks/alignment

src/utils/cropper.py:285

↓ 4 callersMethodcalc_ratio

(self, lmk_lst)

src/live_portrait_wrapper.py:310

↓ 4 callersFunctioncheck_sha1

Check whether the sha1 hash of the file content matches the expected hash. Parameters ---------- filename : str Path to the file.

src/utils/dependencies/insightface/utils/download.py:10

↓ 4 callersMethodcrop_source_image

(self, img_rgb_: np.ndarray, crop_cfg: CropConfig)

src/utils/cropper.py:93

↓ 4 callersFunctionhas_audio_stream

Check if the video file contains an audio stream. :param video_path: Path to the video file :return: True if the video contains an audio

src/utils/video.py:151

↓ 4 callersMethodmake_motion_template

(self, I_lst, c_eyes_lst, c_lip_lst, **kwargs)

src/live_portrait_pipeline.py:41

↓ 4 callersFunctionmkdir

(d, log=False)

src/utils/helper.py:87

↓ 4 callersMethodprepare_source

construct the input as standard img: HxWx3, uint8, 256x256

src/live_portrait_wrapper.py:83

↓ 4 callersMethodtransform_keypoint

transform the implicit keypoints with the pose, shift, and expression deformation kp: BxNx3

src/live_portrait_wrapper.py:184

↓ 3 callersMethod__init__

(self, backbone, position_embedding)

src/utils/dependencies/XPose/models/UniPose/backbone.py:138

↓ 3 callersFunction_run

(command)

src/utils/dependencies/XPose/util/misc.py:349

↓ 3 callersFunction_transform_pts

conduct similarity or affine transformation to the pts pts: Nx2 ndarray M: 2x3 matrix or 3x3 matrix return: Nx2

src/utils/crop.py:36

↓ 3 callersMethodcalc_combined_eye_ratio

(self, c_d_eyes_i, source_lmk)

src/live_portrait_wrapper.py:320

↓ 3 callersFunctioncalc_eye_close_ratio

(lmk: np.ndarray, target_eye_ratio: np.ndarray = None)

src/utils/retargeting_utils.py:14

↓ 3 callersFunctioncalc_lip_close_ratio

(lmk: np.ndarray)

src/utils/retargeting_utils.py:23

↓ 3 callersFunctioncalculate_distance_ratio

(lmk: np.ndarray, idx1: int, idx2: int, idx3: int, idx4: int, eps: float = 1e-6)

src/utils/retargeting_utils.py:9

↓ 3 callersFunctioncrop

(image, target, region)

src/utils/dependencies/XPose/transforms.py:19

↓ 3 callersFunctioncrop_image

(img, pts: np.ndarray, **kwargs)

src/utils/crop.py:376

↓ 3 callersMethodcrop_source_video

Tracking based landmarks/alignment and cropping

src/utils/cropper.py:172

↓ 3 callersMethoddetect

(self, img, input_size = None, max_num=0, metric='default')

src/utils/dependencies/insightface/model_zoo/scrfd.py:220

↓ 3 callersFunctiondownload_file

Download an given URL Parameters ---------- url : str URL to download path : str, optional Destination path to store d

src/utils/dependencies/insightface/utils/download.py:36

↓ 3 callersMethodexecute

(self, args: ArgumentConfig)

src/live_portrait_pipeline.py:77

↓ 3 callersFunctionget_world_size

()

src/utils/dependencies/XPose/util/misc.py:535

↓ 3 callersMethodinference_ctx

(self)

src/live_portrait_wrapper.py:70

↓ 3 callersFunctionis_dist_avail_and_initialized

()

src/utils/dependencies/XPose/util/misc.py:527

↓ 3 callersFunctionis_image

(file_path)

src/utils/helper.py:70

↓ 3 callersFunctionis_square_video

(video_path)

src/utils/helper.py:181

↓ 3 callersFunctionis_video

(file_path)

src/utils/helper.py:75

↓ 3 callersFunctionl2norm

L2-normalize columns of X

src/utils/dependencies/XPose/models/UniPose/fuse_modules.py:40

↓ 3 callersFunctionload_image_rgb

(image_path: str)

src/utils/io.py:12

↓ 3 callersFunctionremove_ddp_dumplicate_key

(state_dict)

src/utils/helper.py:124

↓ 3 callersMethodretarget_eye

kp_source: BxNx3 eye_close_ratio: Bx3 Return: Bx(3*num_kp)

src/live_portrait_wrapper.py:214

↓ 3 callersMethodtic

(self)

src/utils/timer.py:18

↓ 3 callersMethodtoc

(self, average=True)

src/utils/timer.py:23

↓ 3 callersFunctiontrunc_normal_

(tensor, mean=0., std=1., a=-2., b=2.)

src/modules/util.py:441

↓ 3 callersFunctiontry_import

Try import specified package, with custom message support. Parameters ---------- package : str The name of the targeting package.

src/utils/dependencies/insightface/utils/filesystem.py:30

↓ 3 callersMethodwarmup

(self)

src/utils/face_analysis_diy.py:72

↓ 3 callersMethodwith_pos_embed

(tensor, pos)

src/utils/dependencies/XPose/models/UniPose/transformer_deformable.py:381

↓ 3 callersMethodwith_pos_embed

(tensor, pos)

src/utils/dependencies/XPose/models/UniPose/deformable_transformer.py:1047

↓ 2 callersMethod__delattr__

(self, name)

src/utils/dependencies/XPose/util/addict.py:74

↓ 2 callersMethod__init__

(self, input_feat_size, output_feat_size, dropout, do_ln=True)

src/utils/dependencies/XPose/models/UniPose/fuse_modules.py:14

↓ 2 callersMethod__init__

(self, num_pos_feats=64, temperature=10000, normalize=False, scale=None)

src/utils/dependencies/XPose/models/UniPose/position_encoding.py:29

↓ 2 callersMethod_freeze_stages

(self)

src/utils/dependencies/XPose/models/UniPose/swin_transformer.py:553

↓ 2 callersFunction_get_clones

(module, N)

src/utils/dependencies/XPose/models/UniPose/transformer_deformable.py:574

↓ 2 callersMethod_merge_a_into_b

merge dict `a` into dict `b` (non-inplace). values in `a` will overwrite `b`. copy first to avoid inplace modification

src/utils/dependencies/XPose/util/config.py:143

↓ 2 callersMethod_run

(self, inp)

src/utils/human_landmark_runner.py:56

↓ 2 callersMethodactvn

(self, x)

src/modules/util.py:316

next →1–100 of 617, ranked by callers