Functions — github.com/GuyTevet/MotionCLIP

hub / github.com/GuyTevet/MotionCLIP / functions

Functions171 in github.com/GuyTevet/MotionCLIP

⨍ Functions 171◇ Types & classes 10

↓ 7 callersFunctionadding_cuda

(parameters)

src/parser/base.py:19

↓ 7 callersFunctionget_model_and_data

(parameters, split="train")

src/utils/get_model_and_data.py:5

↓ 6 callersMethodrot2xyz

(self, x, mask, get_rotations_back=False, **kwargs)

src/models/modeltype/motionclip.py:56

↓ 5 callersMethod_get_item_data_index

(self, data_index)

src/datasets/dataset.py:144

↓ 5 callersFunctionadd_cuda_options

(parser)

src/parser/base.py:10

↓ 5 callersFunctionload_args

(filename)

src/parser/tools.py:14

↓ 5 callersFunctionload_model_wo_clip

(model, state_dict)

src/utils/misc.py:30

↓ 5 callersFunctionparser

(checkpoint=True)

src/parser/visualize.py:67

↓ 4 callersFunction_copysign

Return a tensor where each element has the absolute value taken from the, corresponding element of a, with sign taken from the corresponding

src/utils/rotation_conversions.py:68

↓ 4 callersFunction_sqrt_positive_part

Returns torch.sqrt(torch.max(0, x)) but with a zero subgradient where x is 0.

src/utils/rotation_conversions.py:86

↓ 4 callersFunctionget_gpu_device

()

src/visualize/visualize.py:425

↓ 4 callersMethodlengths_to_mask

(lengths)

src/models/modeltype/motionclip.py:165

↓ 4 callersFunctionpool_job_with_desc

(pool, iterator, desc, max_, save_path_format, isij)

src/visualize/visualize.py:109

↓ 3 callersFunctionG

(z)

src/models/tools/hessian_penalty.py:126

↓ 3 callersMethod__init__

(self, d_model, dropout=0.1, max_len=5000)

src/models/architectures/transformer.py:29

↓ 3 callersFunction_to_np

(x)

src/visualize/visualize.py:79

↓ 3 callersFunctionadd_dataset_options

(parser)

src/parser/dataset.py:4

↓ 3 callersFunctioncompute_kernel

(x, y)

src/models/tools/mmd.py:5

↓ 3 callersMethodgenerate

(self, classes, durations, nspa=1, # noise_same_action="random", noise_diff_action="random",

src/models/modeltype/motionclip.py:189

↓ 3 callersFunctiongenerate_by_video

(visualization, reconstructions, generation, label_to_action_name, params, nats, nspa, t

src/visualize/visualize.py:63

↓ 3 callersMethodlabel_to_action

(self, label)

src/datasets/dataset.py:45

↓ 3 callersFunctionlistify

If x is already a list, do nothing. Otherwise, wrap x in a list.

src/models/tools/hessian_penalty.py:108

↓ 3 callersFunctionload_anim

(path, timesize=None)

src/visualize/anim.py:43

↓ 3 callersFunctionquaternion_raw_multiply

Multiply two quaternions. Usual torch rules for broadcasting apply. Args: a: Quaternions as tensor of shape (..., 4), real part

src/utils/rotation_conversions.py:340

↓ 3 callersFunctionto_torch

(ndarray)

src/utils/misc.py:13

↓ 3 callersMethodupdate_parameters

(self, parameters)

src/datasets/dataset.py:365

↓ 2 callersFunction_angle_from_tan

Extract the first or third Euler angle from the two members of the matrix which are positive constant times its sine and cosine. Args:

src/utils/rotation_conversions.py:175

↓ 2 callersFunction_index_from_letter

(letter: str)

src/utils/rotation_conversions.py:208

↓ 2 callersFunctionadd_misc_options

(parser)

src/parser/base.py:4

↓ 2 callersFunctionadd_training_options

(parser)

src/parser/training.py:10

↓ 2 callersFunctioncollate_tensors

(batch)

src/utils/tensors.py:10

↓ 2 callersFunctionconstruct_checkpointname

(parameters, folder)

src/parser/checkpoint.py:20

↓ 2 callersFunctionencode_motions

(model, motions, device)

src/visualize/visualize.py:472

↓ 2 callersFunctionget_motion_text_mapping

(datasets)

src/visualize/visualize.py:435

↓ 2 callersFunctionhessian_penalty

Official PyTorch Hessian Penalty implementation. Note: If you want to regularize multiple network activations simultaneously, you need to

src/models/tools/hessian_penalty.py:29

↓ 2 callersFunctionquaternion_to_matrix

Convert rotations given as quaternions to rotation matrices. Args: quaternions: quaternions with real part first, as ten

src/utils/rotation_conversions.py:37

↓ 2 callersFunctionreduction

(x)

src/models/tools/hessian_penalty.py:125

↓ 2 callersFunctionretrieve_motions

(datasets, motion_collection, texts, device)

src/visualize/visualize.py:465

↓ 2 callersFunctionsave_args

(opt, folder)

src/parser/tools.py:5

↓ 2 callersFunctiontrain

(model, optimizer, iterator, device)

src/train/trainer.py:64

↓ 2 callersFunctiontrain_or_test

(model, optimizer, iterator, device, mode="train")

src/train/trainer.py:16

↓ 1 callersMethod_load

(self, ind, frame_ix)

src/datasets/dataset.py:103

↓ 1 callersMethod_load_joints3D

(self, ind, frame_ix)

src/datasets/amass.py:258

↓ 1 callersMethod_load_rotvec

(self, ind, frame_ix)

src/datasets/amass.py:262

↓ 1 callersFunction_test_hessian_penalty

A simple multi-layer test to verify the implementation. Function: G(z) = [z_0 * z_1, z_0**2 * z_1] Ground Truth Hessian Penalty: [4, 16 *

src/models/tools/hessian_penalty.py:116

↓ 1 callersMethodaction_to_action_name

(self, action)

src/datasets/dataset.py:88

↓ 1 callersMethodaction_to_label

(self, action)

src/datasets/dataset.py:42

↓ 1 callersFunctionadd_epochs

(model, datasets, parameters, optimizer, origepoch)

src/train/duration_finetunning.py:14

↓ 1 callersFunctionadd_generation_options

(parser)

src/parser/generate.py:8

↓ 1 callersFunctionadd_model_options

(parser)

src/parser/model.py:5

↓ 1 callersFunctionadd_shadow

(img, shadow=15)

src/visualize/anim.py:35

↓ 1 callersFunctionadd_visualize_options

(parser)

src/parser/visualize.py:14

↓ 1 callersFunctionaxis_angle_to_quaternion

Convert rotations given as axis/angle to quaternions. Args: axis_angle: Rotations given as a vector in axis angle form,

src/utils/rotation_conversions.py:449

↓ 1 callersFunctioncollate

(batch)

src/utils/tensors.py:23

↓ 1 callersMethodcompute_clip_losses

(self, batch)

src/models/modeltype/motionclip.py:85

↓ 1 callersMethodcompute_loss

(self, batch)

src/models/modeltype/motionclip.py:61

↓ 1 callersFunctioncompute_mmd

(x, y)

src/models/tools/mmd.py:17

↓ 1 callersFunctionconstruct_figname

(parameters)

src/parser/visualize.py:9

↓ 1 callersFunctionconvert_models_to_fp32

(model)

src/train/trainer.py:9

↓ 1 callersFunctiondo_epochs

(model, datasets, parameters, optimizer, writer)

src/train/train.py:17

↓ 1 callersFunctionfixseed

(seed)

src/utils/fixseed.py:6

↓ 1 callersMethodget_action

(self, ind)

src/datasets/dataset.py:85

↓ 1 callersFunctionget_babel_labels

(babel_dir_path)

src/datasets/amass_parser.py:265

↓ 1 callersMethodget_clip_action_cat

(self, ind, frame_ix)

src/datasets/dataset.py:70

↓ 1 callersMethodget_clip_image

(self, ind)

src/datasets/dataset.py:58

↓ 1 callersMethodget_clip_path

(self, ind)

src/datasets/dataset.py:62

↓ 1 callersMethodget_clip_text

(self, ind, frame_ix)

src/datasets/dataset.py:66

↓ 1 callersFunctionget_datasets

(parameters, clip_preprocess, split="train")

src/datasets/get_dataset.py:7

↓ 1 callersFunctionget_joints_to_use

(args)

src/datasets/amass_parser.py:41

↓ 1 callersMethodget_label

(self, ind)

src/datasets/dataset.py:74

↓ 1 callersMethodget_label_sample

(self, label, n=1, return_labels=False, return_index=False)

src/datasets/dataset.py:251

↓ 1 callersFunctionget_loss_function

(ltype)

src/models/tools/losses.py:84

↓ 1 callersMethodget_mean_length_label

(self, label)

src/datasets/dataset.py:311

↓ 1 callersFunctionget_palette

(i, nspa)

src/visualize/visualize.py:90

↓ 1 callersMethodget_pose_data

(self, data_index, frame_ix)

src/datasets/dataset.py:53

↓ 1 callersFunctionget_z

Solves for the depth offset of the model to approx. orth with persp camera.

src/datasets/amass.py:99

↓ 1 callersFunctionlengths_to_mask

(lengths)

src/utils/tensors.py:4

↓ 1 callersMethodload_db

(self)

src/datasets/amass.py:242

↓ 1 callersFunctionmain

()

src/visualize/text2motion.py:16

↓ 1 callersFunctionmain

()

src/visualize/motion_interpolation.py:19

↓ 1 callersFunctionmain

()

src/visualize/motion_editing.py:18

↓ 1 callersFunctionmain

()

src/train/duration_finetunning.py:34

↓ 1 callersFunctionmatrix_to_quaternion

Convert rotations given as rotation matrices to quaternions. Args: matrix: Rotation matrices as tensor of shape (..., 3, 3). Re

src/utils/rotation_conversions.py:97

↓ 1 callersFunctionmulti_layer_second_directional_derivative

Estimates the second directional derivative of G w.r.t. its input at z in the direction x

src/models/tools/hessian_penalty.py:75

↓ 1 callersFunctionmulti_stack_var_and_reduce

Iterate over all activations to be regularized, then apply Equation (5) to each.

src/models/tools/hessian_penalty.py:99

↓ 1 callersFunctionparse_info_name

(path)

src/datasets/tools.py:5

↓ 1 callersFunctionparse_modelname

(modelname)

src/parser/model.py:38

↓ 1 callersFunctionparser

()

src/parser/checkpoint.py:6

↓ 1 callersFunctionparser

()

src/parser/training.py:18

↓ 1 callersFunctionplot_3d_motion

(motion, length, save_path, params, title="", interval=50, palette=None)

src/visualize/anim.py:57

↓ 1 callersFunctionquaternion_invert

Given a quaternion representing rotation, get the quaternion representing its inverse. Args: quaternion: Quaternions as tensor o

src/utils/rotation_conversions.py:378

↓ 1 callersFunctionquaternion_to_axis_angle

Convert rotations given as quaternions to axis/angle. Args: quaternions: quaternions with real part first, as tensor of

src/utils/rotation_conversions.py:481

↓ 1 callersFunctionrademacher

Creates a random tensor of size [shape] under the Rademacher distribution (P(x=1) == P(x=-1) == 0.5)

src/models/tools/hessian_penalty.py:67

↓ 1 callersFunctionrandom_quaternions

Generate random quaternions representing rotations, i.e. versors with nonnegative real part. Args: n: Number of quaternions in a

src/utils/rotation_conversions.py:259

↓ 1 callersFunctionrandom_rotations

Generate random rotations as 3x3 rotation matrices. Args: n: Number of rotation matrices in a batch to return. dtype: Type t

src/utils/rotation_conversions.py:283

↓ 1 callersFunctionread_data

(folder, split_name,dataset_name, target_fps, max_fps_dist, joints_to_use, quick_run,babel_labels, clip_images

src/datasets/amass_parser.py:80

↓ 1 callersFunctionread_single_sequence

(split_name, dataset_name, folder, seq_name, body_models, target_fps, max_fps_dist, j

src/datasets/amass_parser.py:123

↓ 1 callersFunctionstack_gen_and_images

(gen, images)

src/visualize/visualize.py:35

↓ 1 callersFunctionstack_gen_only

(gen)

src/visualize/visualize.py:48

↓ 1 callersFunctionstack_images

(real, real_gens, gen)

src/visualize/visualize.py:22

next →1–100 of 171, ranked by callers