Functions — github.com/TencentARC/Pixal3D

hub / github.com/TencentARC/Pixal3D / functions

Functions865 in github.com/TencentARC/Pixal3D

⨍ Functions 865◇ Types & classes 147↳ Endpoints 3

↓ 1 callersMethod__init__

( self, *args, t_schedule: dict = { 'name': 'logitNormal', 'ar

pixal3d/trainers/flow_matching/flow_matching.py:53

↓ 1 callersMethod__init__

(self, *args, image_cond_model: dict, **kwargs)

pixal3d/trainers/flow_matching/mixins/image_conditioned.py:130

↓ 1 callersMethod__init__

(self, roots: str, *, min_aesthetic_score: float = 5.0, max_tokens: int = 3276

pixal3d/datasets/structured_latent.py:112

↓ 1 callersMethod__init__

(self, hidden_size, frequency_embedding_size=256)

pixal3d/models/sparse_structure_flow.py:16

↓ 1 callersMethod__init__

( self, model_channels: List[int], latent_channels: int, num_blocks: List[int]

pixal3d/models/sc_vaes/fdg_vae.py:24

↓ 1 callersMethod__merge_sparse_cache

(self, other: 'SparseTensor')

pixal3d/modules/sparse/basic.py:705

↓ 1 callersMethod_add_sample

(self, memory, input_size, mem_ratio)

pixal3d/utils/elastic_utils.py:79

↓ 1 callersMethod_delete_slat_dec

(self)

pixal3d/datasets/structured_latent_svpbr.py:62

↓ 1 callersMethod_delete_slat_dec

(self)

pixal3d/datasets/structured_latent.py:41

↓ 1 callersMethod_delete_ss_dec

(self)

pixal3d/datasets/sparse_structure_latent.py:42

↓ 1 callersFunction_dict_merge

Merge two dictionaries.

pixal3d/utils/general_utils.py:9

↓ 1 callersMethod_extract_dino_features

Extract DINOv3 features from normalized image.

pixal3d/trainers/flow_matching/mixins/image_conditioned_proj.py:711

↓ 1 callersMethod_extract_vae_latent

Extract Flux VAE latent from unnormalized image [0,1].

pixal3d/trainers/flow_matching/mixins/image_conditioned_proj.py:724

↓ 1 callersMethod_fit_params

(self)

pixal3d/utils/elastic_utils.py:100

↓ 1 callersMethod_forward

(self, x: SparseTensor)

pixal3d/modules/sparse/transformer/blocks.py:63

↓ 1 callersMethod_forward

(self, x: SparseTensor, context: Union[torch.Tensor, VarLenTensor])

pixal3d/modules/sparse/transformer/blocks.py:129

↓ 1 callersMethod_forward

(self, x: SparseTensor, mod: torch.Tensor)

pixal3d/modules/sparse/transformer/modulated.py:57

↓ 1 callersMethod_forward

(self, x: SparseTensor, mod: torch.Tensor, context: Union[torch.Tensor, VarLenTensor])

pixal3d/modules/sparse/transformer/modulated.py:181

↓ 1 callersMethod_forward

(self, x: torch.Tensor, phases: Optional[torch.Tensor] = None)

pixal3d/modules/transformer/blocks.py:101

↓ 1 callersMethod_forward

(self, x: torch.Tensor, context: torch.Tensor, phases: Optional[torch.Tensor] = None)

pixal3d/modules/transformer/blocks.py:169

↓ 1 callersMethod_forward

(self, x: torch.Tensor, mod: torch.Tensor, phases: Optional[torch.Tensor] = None)

pixal3d/modules/transformer/modulated.py:56

↓ 1 callersMethod_forward

(self, x: torch.Tensor, mod: torch.Tensor, context: torch.Tensor, phases: Optional[torch.Tensor] = None)

pixal3d/modules/transformer/modulated.py:180

↓ 1 callersMethod_forward

(self, x: sp.SparseTensor)

pixal3d/models/sc_vaes/sparse_unet_vae.py:66

↓ 1 callersMethod_forward

(self, x: sp.SparseTensor)

pixal3d/models/sc_vaes/sparse_unet_vae.py:112

↓ 1 callersMethod_forward

(self, x: sp.SparseTensor, subdiv: sp.SparseTensor = None)

pixal3d/models/sc_vaes/sparse_unet_vae.py:154

↓ 1 callersMethod_forward

(self, x: sp.SparseTensor)

pixal3d/models/sc_vaes/sparse_unet_vae.py:198

↓ 1 callersMethod_forward

(self, x: sp.SparseTensor, subdiv: sp.SparseTensor = None)

pixal3d/models/sc_vaes/sparse_unet_vae.py:240

↓ 1 callersMethod_forward

(self, x: sp.SparseTensor)

pixal3d/models/sc_vaes/sparse_unet_vae.py:284

↓ 1 callersMethod_get_input_size

Get the size of the input data. Returns: int: The size of the input data.

pixal3d/utils/elastic_utils.py:157

↓ 1 callersMethod_get_input_size

Get the size of the input data. Returns: int: The size of the input data.

pixal3d/utils/elastic_utils.py:196

↓ 1 callersMethod_get_model_prediction

(self, model, x_t, t, cond=None, **kwargs)

pixal3d/pipelines/samplers/flow_euler.py:48

↓ 1 callersMethod_get_phases

(self, indices: torch.Tensor)

pixal3d/modules/attention/rope.py:22

↓ 1 callersMethod_get_phases

(self, indices: torch.Tensor)

pixal3d/modules/sparse/attention/rope.py:23

↓ 1 callersMethod_inference_model

(self, model, x_t, t, cond=None, **kwargs)

pixal3d/pipelines/samplers/flow_euler.py:44

↓ 1 callersMethod_init_image_cond_model

Initialize the image conditioning model.

pixal3d/trainers/flow_matching/mixins/image_conditioned.py:135

↓ 1 callersMethod_init_image_cond_model

Initialize the image conditioning model.

pixal3d/trainers/flow_matching/mixins/image_conditioned.py:190

↓ 1 callersMethod_init_text_cond_model

Initialize the text conditioning model.

pixal3d/trainers/flow_matching/mixins/text_conditioned.py:22

↓ 1 callersFunction_install_blender

()

data_toolkit/dump_pbr.py:19

↓ 1 callersFunction_install_blender

()

data_toolkit/dump_mesh.py:19

↓ 1 callersFunction_install_blender

()

data_toolkit/render_cond.py:21

↓ 1 callersMethod_load_vae

Lazy-load Flux VAE encoder.

pixal3d/trainers/flow_matching/mixins/image_conditioned_proj.py:671

↓ 1 callersMethod_loading_slat_dec

(self)

pixal3d/datasets/structured_latent_svpbr.py:40

↓ 1 callersMethod_loading_slat_dec

(self)

pixal3d/datasets/structured_latent.py:29

↓ 1 callersMethod_loading_ss_dec

(self)

pixal3d/datasets/sparse_structure_latent.py:30

↓ 1 callersFunction_naive_sdpa

Naive implementation of scaled dot product attention.

pixal3d/modules/attention/full_attn.py:12

↓ 1 callersMethod_run_projection_visualization_test

Run projection visualization test on a few samples before training starts. This helps verify that the 3D-to-2D projection is

pixal3d/trainers/flow_matching/sparse_flow_matching.py:572

↓ 1 callersMethod_run_projection_visualization_test

Run projection visualization test on a few samples before training starts. This helps verify that the 3D-to-2D projection is

pixal3d/trainers/flow_matching/flow_matching.py:612

↓ 1 callersMethod_sin_cos_embedding

Create sinusoidal position embeddings. Args: x: a 1-D Tensor of N indices Returns: an (N, D) Tensor

pixal3d/modules/transformer/blocks.py:20

↓ 1 callersFunction_ssim

(img1, img2, window, window_size, channel, size_average=True)

pixal3d/utils/loss_utils.py:49

↓ 1 callersMethod_v_to_xstart_eps

(self, x_t, t, v)

pixal3d/pipelines/samplers/flow_euler.py:32

↓ 1 callersMethod_verify_gradient_sync

Verify that DDP gradient synchronization is working correctly. DDP's backward automatically performs all_reduce on gradients; after s

pixal3d/trainers/basic.py:904

↓ 1 callersMethod_xstart_to_pred

(self, x_t, t, x_0)

pixal3d/pipelines/samplers/flow_euler.py:41

↓ 1 callersFunctionbuild_downloaded_metadata_from_files

Scan local files under raw_root to build download metadata. Walks through raw_root to find downloaded 3D files (.glb, .obj, .fbx, .usdz, .glt

data_toolkit/build_metadata.py:54

↓ 1 callersMethodcheck_abort

Check if training should be aborted due to certain conditions.

pixal3d/trainers/basic.py:1149

↓ 1 callersFunctioncheck_mask_boundary_distance

Check the rendered object's mask distance to image boundary. Args: image_path: Path to the rendered PNG image with alpha channel.

data_toolkit/blender_script/render_cond.py:367

↓ 1 callersMethodclear_spatial_cache

Clear all spatial caches.

pixal3d/modules/sparse/basic.py:767

↓ 1 callersFunctioncompute_f_pixels

(camera_angle_x: float, resolution: int)

app.py:208

↓ 1 callersFunctioncompute_f_pixels

(camera_angle_x: float, resolution: int)

inference.py:118

↓ 1 callersMethodconvert_to

Convert the torso of the model to the specified dtype.

pixal3d/models/sparse_structure_flow.py:165

↓ 1 callersMethodconvert_to

Convert the torso of the model to the specified dtype.

pixal3d/models/structured_latent_flow.py:111

↓ 1 callersMethodconvert_to_fp16

Convert the torso of the model to float16.

pixal3d/models/sparse_structure_vae.py:168

↓ 1 callersMethodconvert_to_fp16

Convert the torso of the model to float16.

pixal3d/models/sparse_structure_vae.py:277

↓ 1 callersMethodconvert_to_fp16

Convert the torso of the model to float16.

pixal3d/models/sc_vaes/sparse_unet_vae.py:352

↓ 1 callersMethodconvert_to_fp16

Convert the torso of the model to float16.

pixal3d/models/sc_vaes/sparse_unet_vae.py:457

↓ 1 callersMethodcpu

(self)

pixal3d/modules/image_feature_extractor.py:28

↓ 1 callersMethodcpu

(self)

pixal3d/pipelines/rembg/BiRefNet.py:28

↓ 1 callersFunctioncreate_window

(window_size, channel)

pixal3d/utils/loss_utils.py:27

↓ 1 callersFunctioncube_to_dir

(s, x, y)

pixal3d/renderers/pbr_mesh_renderer.py:10

↓ 1 callersMethoddecode_latent

Decode the latent codes. Args: shape_slat (SparseTensor): The structured latent for shape. tex_slat (SparseT

pixal3d/pipelines/trellis2_image_to_3d.py:467

↓ 1 callersMethoddecode_latent

Decode the latent codes. Args: shape_slat (SparseTensor): The structured latent for shape. tex_slat (SparseT

pixal3d/pipelines/pixal3d_image_to_3d.py:575

↓ 1 callersMethoddecode_latent

(self, z, shape_z, batch_size=4)

pixal3d/datasets/structured_latent_svpbr.py:69

↓ 1 callersMethoddecode_latent

(self, z, batch_size=4)

pixal3d/datasets/sparse_structure_latent.py:47

↓ 1 callersMethoddecode_shape_slat

Decode the structured latent. Args: slat (SparseTensor): The structured latent. Returns: List[Mesh]

pixal3d/pipelines/trellis2_image_to_3d.py:377

↓ 1 callersMethoddecode_shape_slat

Decode the structured latent. Args: slat (SparseTensor): The structured latent. Returns: List[Mesh]

pixal3d/pipelines/pixal3d_image_to_3d.py:485

↓ 1 callersMethoddecode_tex_slat

Decode the structured latent. Args: slat (SparseTensor): The structured latent. Returns: SparseTens

pixal3d/pipelines/trellis2_image_to_3d.py:445

↓ 1 callersMethoddecode_tex_slat

Decode the structured latent. Args: slat (SparseTensor): The structured latent. Returns: SparseTens

pixal3d/pipelines/pixal3d_image_to_3d.py:553

↓ 1 callersMethoddecode_tex_slat

Decode the structured latent. Args: slat (SparseTensor): The structured latent. Returns: SparseTens

pixal3d/pipelines/trellis2_texturing.py:267

↓ 1 callersFunctiondelete_invisible_objects

Deletes all invisible objects in the scene. Returns: None

data_toolkit/blender_script/dump_pbr.py:93

↓ 1 callersFunctiondelete_invisible_objects

Deletes all invisible objects in the scene. Returns: None

data_toolkit/blender_script/dump_mesh.py:92

↓ 1 callersFunctiondelete_invisible_objects

Deletes all invisible objects in the scene. Returns: None

data_toolkit/blender_script/render_cond.py:226

↓ 1 callersFunctiondict_any

Recursively apply a function to all non-dictionary leaf values in a dictionary.

pixal3d/utils/general_utils.py:74

↓ 1 callersFunctiondict_flatten

Flatten a nested dictionary into a dictionary with no nested dictionaries.

pixal3d/utils/general_utils.py:104

↓ 1 callersMethodencode_shape_slat

Encode the meshes to structured latent. Args: mesh (trimesh.Trimesh): The mesh to encode. resolution (int):

pixal3d/pipelines/trellis2_texturing.py:183

↓ 1 callersMethodextract_features

(self, image: torch.Tensor)

pixal3d/modules/image_feature_extractor.py:81

↓ 1 callersMethodextract_features

Extract features using DINOv3.

pixal3d/trainers/flow_matching/mixins/image_conditioned_proj.py:450

↓ 1 callersMethodextract_features

(self, image: torch.Tensor)

pixal3d/trainers/flow_matching/mixins/image_conditioned.py:83

↓ 1 callersMethodfinetune_from

Finetune from a checkpoint. Should be called by all processes.

pixal3d/trainers/basic.py:489

↓ 1 callersMethodforward

(self, vertices: sp.SparseTensor, intersected: sp.SparseTensor, sample_posterior=False, return_raw=False)

pixal3d/models/sc_vaes/fdg_vae.py:45

↓ 1 callersFunctiongaussian

(window_size, sigma)

pixal3d/utils/loss_utils.py:22

↓ 1 callersFunctionget_camera_params_wild_moge

(image_path, device="cuda", mesh_scale=1.0, extend_pixel=0, image_resolution=512)

app.py:225

↓ 1 callersFunctionget_camera_params_wild_moge

(image_path, moge_model, device="cuda", mesh_scale=1.0, extend_pixel=0, image_resolution=512)

inference.py:137

↓ 1 callersMethodget_cond

Get the conditioning data.

pixal3d/trainers/flow_matching/mixins/image_conditioned.py:153

↓ 1 callersFunctionget_file_hash

(file: str)

data_toolkit/utils.py:7

↓ 1 callersFunctionget_file_hash

(file: str)

data_toolkit/build_texverse_metadata.py:24

↓ 1 callersMethodget_inference_cond

Get the conditioning data for inference.

pixal3d/trainers/flow_matching/mixins/image_conditioned.py:162

↓ 1 callersFunctionget_model_summary

(model)

train.py:49

↓ 1 callersMethodget_proj_cond_ss

Get proj conditioning for sparse structure stage. Args: image: List of PIL images. camera_angle_x: Camera ho

pixal3d/pipelines/pixal3d_image_to_3d.py:193

↓ 1 callersFunctionget_transform_matrix

(obj: bpy.types.Object)

data_toolkit/blender_script/render_cond.py:353

↓ 1 callersFunctionhalton_sequence

(dim, n)

data_toolkit/utils.py:31

↓ 1 callersFunctionhalton_sequence

(dim, n)

pixal3d/utils/random_utils.py:16

← previous next →201–300 of 865, ranked by callers