Functions — github.com/CUT3R/CUT3R

hub / github.com/CUT3R/CUT3R / functions

Functions1,487 in github.com/CUT3R/CUT3R

⨍ Functions 1,487◇ Types & classes 174

↓ 164 callersFunctionprint

(*args, **kwargs)

src/croco/utils/misc.py:207

↓ 146 callersFunctionprint

(*args, **kwargs)

src/train.py:68

↓ 62 callersMethodsize

(self)

src/dust3r/datasets/utils/cropping.py:45

↓ 55 callersFunctionimread_cv2

Open an image or a depthmap with opencv-python.

src/dust3r/utils/image.py:34

↓ 32 callersMethod_crop_resize_if_necessary

This function: - first downsizes the image with LANCZOS inteprolation, which is better than bilinear interpolation in

src/dust3r/datasets/base/base_multiview_dataset.py:443

↓ 32 callersFunctiongeotrf

Apply a geometric transformation to a list of 3-D points. H: 3x3 or 4x4 projection matrix (typically a Homography) p: numpy/torch/tuple of co

src/dust3r/utils/geometry.py:49

↓ 32 callersMethodupdate

(self, **kwargs)

src/croco/utils/misc.py:96

↓ 30 callersMethodmax

(self)

src/croco/utils/misc.py:74

↓ 26 callersFunctionadd_arg

( name_or_flags, default=None, default_stereo=None, default_flow=None, **kwargs )

src/croco/stereoflow/train.py:51

↓ 23 callersMethodmedian

(self)

src/croco/utils/misc.py:62

↓ 22 callersMethodget_seq_from_start_id

args: num_views: number of views to return id_ref: the reference id (first id) ids_all: all the ids

src/dust3r/datasets/base/base_multiview_dataset.py:178

↓ 19 callersMethodget_img_and_ray_masks

(self, is_metric, v, rng, p=[0.8, 0.15, 0.05])

src/dust3r/datasets/base/base_multiview_dataset.py:291

↓ 19 callersFunctionwrite_next_bytes

pack and write to a binary file. :param fid: :param data: data to send, if multiple elements are sent at the same time, they should be enc

datasets_preprocess/read_write_model.py:88

↓ 17 callersMethodresize

(self, *args, **kwargs)

src/dust3r/datasets/utils/cropping.py:50

↓ 16 callersFunctioninv

Invert a torch or numpy matrix

src/dust3r/utils/geometry.py:118

↓ 16 callersFunctionto_numpy

(x)

src/dust3r/utils/device.py:42

↓ 15 callersFunctioninvalid_to_nans

(arr, valid_mask, ndim=999)

src/dust3r/utils/misc.py:109

↓ 13 callersFunctionread_next_bytes

Read and unpack the next bytes from a binary file. :param fid: :param num_bytes: Sum of combination of {2, 4, 8}, e.g. 2, 6, 16, 30, etc.

datasets_preprocess/read_write_model.py:76

↓ 12 callersMethod_set_root

(self)

src/croco/stereoflow/datasets_stereo.py:138

↓ 11 callersMethodresize

(self, *args, **kwargs)

eval/mv_recon/dataset_utils/cropping.py:49

↓ 10 callersFunctionbatch_append

(original_list, new_list)

src/train.py:646

↓ 10 callersFunctionedge_str

(i, j)

cloud_opt/utils.py:79

↓ 10 callersFunctionedge_str

(i, j)

cloud_opt/dust3r_opt/commons.py:12

↓ 9 callersFunctionParameterStack

(params, keys=None, is_param=None, fill=0)

cloud_opt/dust3r_opt/optimizer.py:263

↓ 9 callersMethod__init__

( self, in_features, hidden_features=None, out_features=None, act_laye

src/dust3r/blocks.py:62

↓ 9 callersMethod__init__

(self)

src/dust3r/losses.py:167

↓ 9 callersFunctiondepth_evaluation

Evaluate the depth map using various metrics and return a depth error parity map, with an option for least squares alignment. Args:

eval/video_depth/tools.py:126

↓ 8 callersMethodcrop

(self, *args, **kwargs)

src/dust3r/datasets/utils/cropping.py:53

↓ 8 callersMethodresize

(self, *args, **kwargs)

datasets_preprocess/utils/cropping.py:72

↓ 7 callersMethod__init__

(self, max_gtnorm=None)

src/croco/stereoflow/criterion.py:25

↓ 7 callersMethod_set_root

(self)

src/croco/stereoflow/datasets_flow.py:120

↓ 7 callersFunctionmake_traj

(args)

eval/relpose/evo_utils.py:151

↓ 7 callersFunctionpostprocess

extract 3D points/confidence from prediction head output

src/dust3r/heads/postprocess.py:11

↓ 7 callersMethodstate_dict

(self)

src/croco/utils/misc.py:311

↓ 6 callersMethod__init__

( self, in_features, hidden_features=None, out_features=None, act_laye

src/croco/models/blocks.py:71

↓ 6 callersMethod_crop_resize_if_necessary

This function: - first downsizes the image with LANCZOS inteprolation, which is better than bilinear interpolation in

eval/mv_recon/base.py:167

↓ 6 callersFunction_resize_pil_image

(img, long_edge_size)

src/dust3r/utils/image.py:65

↓ 6 callersFunctionadd_path_to_dust3r

(ckpt)

add_ckpt_path.py:6

↓ 6 callersMethodbackward

(ctx, grad_res)

src/croco/models/curope/curope2d.py:25

↓ 6 callersMethodclose

(self)

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:250

↓ 6 callersFunctioncolorize

turn a grayscale image into a color image :param x: torch.Tensor, grayscale image, [H, W] or [B, H, W] :param mask: torch.Tensor or None,

src/dust3r/viz.py:567

↓ 6 callersMethoddevice

(self)

cloud_opt/dust3r_opt/base_opt.py:197

↓ 6 callersMethodfrom_pretrained

(cls, pretrained_model_name_or_path, **kw)

src/dust3r/model.py:306

↓ 6 callersFunctiongeotrf

Apply a geometric transformation to a list of 3-D points. H: 3x3 or 4x4 projection matrix (typically a Homography) p: numpy/torch/tuple of co

cloud_opt/utils.py:264

↓ 6 callersMethodget_im_poses

(self)

cloud_opt/dust3r_opt/base_opt.py:315

↓ 6 callersMethodget_pts3d

(self, raw=False)

cloud_opt/dust3r_opt/base_opt.py:293

↓ 6 callersFunctiongroup_by_directory

Groups the file paths based on the second-to-last directory in their paths. Parameters: - pathes (list): List of file paths. Return

eval/video_depth/tools.py:14

↓ 6 callersMethodload_state_dict

(self, state_dict)

src/croco/utils/misc.py:317

↓ 6 callersFunctionsRT_to_4x4

(scale, R, T, device)

cloud_opt/utils.py:257

↓ 6 callersFunctionsRT_to_4x4

(scale, R, T, device)

cloud_opt/dust3r_opt/init_im_poses.py:274

↓ 6 callersMethodset_epoch

(self, epoch)

src/dust3r/datasets/base/easy_dataset.py:166

↓ 5 callersMethodadd_pointcloud

(self, pts3d, color=(0, 0, 0), mask=None, denoise=False)

src/dust3r/viz.py:165

↓ 5 callersFunctiondepth2disparity

(depth, return_mask=False)

eval/video_depth/tools.py:34

↓ 5 callersFunctiondepth2disparity

(depth, return_mask=False)

eval/monodepth/tools.py:34

↓ 5 callersFunctionestimate_focal_knowing_depth

Reprojection method, for when the absolute depth is known: 1) estimate the camera focal using a robust estimator 2) reproject points onto true

src/dust3r/post_process.py:12

↓ 5 callersMethodget_depthmaps

(self, raw=False)

cloud_opt/dust3r_opt/base_opt.py:321

↓ 5 callersFunctioninference

(groups, model, device, verbose=True)

src/dust3r/inference.py:220

↓ 5 callersFunctioninvalid_to_zeros

(arr, valid_mask, ndim=999)

src/dust3r/utils/misc.py:118

↓ 5 callersFunctionload_images

open and convert all images in a list or folder to proper input format for DUSt3R

src/dust3r/utils/image.py:75

↓ 5 callersMethodlog_every

( self, iterable, print_freq, accelerator: Accelerator, header=None, max_iter=None )

src/croco/utils/misc.py:131

↓ 5 callersMethodnormalize

(self, x, mask)

src/dust3r/losses.py:239

↓ 5 callersFunctionpose_encoding_to_camera

Args: pose_encoding: A tensor of shape `BxC`, containing a batch of `B` `C`-dimensional pose encodings. p

src/dust3r/utils/camera.py:396

↓ 5 callersFunctionprocess_scannet

(args, img_path)

eval/relpose/metadata.py:199

↓ 5 callersFunctionreg_dense_conf

extract confidence from prediction head output

src/dust3r/heads/postprocess.py:142

↓ 5 callersMethodrender_viewpoint

(self, viewpoint_position, viewpoint_orientation)

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:332

↓ 5 callersFunctionto_numpy

(x)

viser_utils.py:56

↓ 4 callersFunctionNoGradParamDict

(x)

cloud_opt/utils.py:74

↓ 4 callersFunctionNoGradParamDict

(x)

cloud_opt/dust3r_opt/commons.py:28

↓ 4 callersMethod__init__

(self, criterion, norm_mode="avg_dis", gt_scale=False, fix_first=True)

eval/mv_recon/criterion.py:277

↓ 4 callersFunction_check_input

(value, center=1, bound=(0, float("inf")), clip_first_on_zero=True)

eval/mv_recon/dataset_utils/transforms.py:13

↓ 4 callersFunction_check_input

(value, center=1, bound=(0, float("inf")), clip_first_on_zero=True)

src/dust3r/datasets/utils/transforms.py:14

↓ 4 callersMethod_downstream_head

(self, decout, img_shape, **kwargs)

src/dust3r/model.py:700

↓ 4 callersMethod_encode_ray_map

(self, ray_map, true_shape)

src/dust3r/model.py:527

↓ 4 callersFunction_get_gtnorm

(gt)

src/croco/stereoflow/criterion.py:13

↓ 4 callersMethod_get_img_level_feat

(self, feat)

src/dust3r/model.py:732

↓ 4 callersMethod_get_msk_indices

(self, msk)

cloud_opt/dust3r_opt/optimizer.py:138

↓ 4 callersFunction_read_img

(filename)

src/croco/stereoflow/datasets_stereo.py:826

↓ 4 callersMethod_recurrent_rollout

( self, state_feat, state_pos, current_feat, current_pos, pose

src/dust3r/model.py:713

↓ 4 callersMethod_set_pose

(self, poses, idx, R, T=None, scale=None, force=False)

cloud_opt/dust3r_opt/base_opt.py:243

↓ 4 callersFunctioncollate_with_cat

(whatever, lists=False)

demo_ga.py:50

↓ 4 callersFunctioncompute_pointcloud

Return a 3D point cloud corresponding to valid pixels of the depth map

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:45

↓ 4 callersFunctiondepth_evaluation

Evaluate the depth map using various metrics and return a depth error parity map, with an option for least squares alignment. Args:

eval/monodepth/tools.py:126

↓ 4 callersFunctionfloat2uint8

(x)

src/dust3r/viz.py:31

↓ 4 callersMethodget_all_pts3d_t

(self, gts, preds, dist_clip=None)

eval/mv_recon/criterion.py:283

↓ 4 callersMethodget_focals

(self)

cloud_opt/dust3r_opt/base_opt.py:302

↓ 4 callersFunctionget_group_pointcloud_center_scale

(pts, valid_masks=None, z_only=False, center=True)

src/dust3r/utils/geometry.py:474

↓ 4 callersMethodget_norm_factor_point_cloud

( self, pts_self, pts_cross, valids, conf_self, conf_cross, norm_self_only=False )

src/dust3r/losses.py:319

↓ 4 callersMethodget_norm_factor_poses

(self, gt_trans, pr_trans, not_metric_mask)

src/dust3r/losses.py:335

↓ 4 callersMethodget_pw_poses

(self)

cloud_opt/dust3r_opt/base_opt.py:279

↓ 4 callersFunctionget_render_results

(gts, preds, self_view=False)

src/dust3r/utils/render.py:54

↓ 4 callersFunctionimg_to_tensor

(img)

src/croco/stereoflow/datasets_stereo.py:45

↓ 4 callersMethodinit

(self, dim_tokens_enc=768)

src/dust3r/heads/dpt_head.py:32

↓ 4 callersMethodinquire

(self, query, mem)

src/dust3r/model.py:217

↓ 4 callersFunctionlist_scenes

Generic method iterating through a base_path folder to find scenes.

src/croco/datasets/habitat_sim/paths.py:97

↓ 4 callersFunctionmake_fusion_block

(features, use_bn, width_ratio=1)

src/croco/models/dpt_block.py:235

↓ 4 callersFunctionmy_formatting

(val)

src/croco/datasets/habitat_sim/generate_multiview_images.py:165

↓ 4 callersFunctionnormalize_pointcloud_group

renorm pointmaps pts1, pts2 with norm_mode

src/dust3r/utils/geometry.py:329

↓ 4 callersFunctionpostprocess_pose

extract pose from prediction head output

src/dust3r/heads/postprocess.py:30

↓ 4 callersFunctionpostprocess_rgb

(out, eps=1e-6)

src/dust3r/heads/postprocess.py:23

↓ 4 callersFunctionsave_model

(epoch, fname, best_so_far)

src/train.py:247

next →1–100 of 1,487, ranked by callers