Functions — github.com/Tsinghua-MARS-Lab/SLAM-Former

hub / github.com/Tsinghua-MARS-Lab/SLAM-Former / functions

Functions601 in github.com/Tsinghua-MARS-Lab/SLAM-Former

⨍ Functions 601◇ Types & classes 107

↓ 168 callersFunctionprint

(*args, **kwargs)

src/croco/utils/misc.py:209

↓ 26 callersFunctionadd_arg

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

src/croco/stereoflow/train.py:51

↓ 17 callersMethodtime

(self)

slam/demo.py:145

↓ 15 callersMethodupdate

(self, **kwargs)

src/croco/utils/misc.py:96

↓ 13 callersMethodmax

(self)

src/croco/utils/misc.py:74

↓ 12 callersMethod_set_root

(self)

src/croco/stereoflow/datasets_stereo.py:138

↓ 9 callersMethodload_state_dict

(self, state_dict)

src/croco/utils/misc.py:322

↓ 8 callersFunction_make_dinov2_model

( *, arch_name: str = "vit_large", img_size: int = 518, patch_size: int = 14, init_values:

src/slamformer/models/dinov2/hub/backbones.py:18

↓ 8 callersMethoddecode

during inference there is only branch1 and branch2 because branch3==branch1 now with just different cache

src/slamformer/models/slamformer.py:238

↓ 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 callersFunctionlog_camera

( entity_path: str, world_T_cam_44: torch.Tensor, K_44: torch.Tensor, kfd=False, update=False, )

slam/rerun_helper/__init__.py:52

↓ 6 callersMethod__init__

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

src/croco/models/blocks.py:71

↓ 6 callersMethodextract

(self, map_all=None)

slam/demo.py:425

↓ 5 callersMethodrender_viewpoint

(self, viewpoint_position, viewpoint_orientation)

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

↓ 4 callersFunction_get_gtnorm

(gt)

src/croco/stereoflow/criterion.py:13

↓ 4 callersFunction_read_img

(filename)

src/croco/stereoflow/datasets_stereo.py:826

↓ 4 callersFunctioncolormap_image

Colormaps a one channel tensor using a matplotlib colormap. Args: image_1hw: the tensor to colomap. mask_1hw: an optional f

slam/rerun_helper/visualization_utils.py:12

↓ 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 callersFunctiondrop_add_residual_stochastic_depth

( x: Tensor, residual_func: Callable[[Tensor], Tensor], sample_drop_ratio: float = 0.0, )

src/slamformer/models/layers/block.py:115

↓ 4 callersFunctionimg_to_tensor

(img)

src/croco/stereoflow/datasets_stereo.py:45

↓ 4 callersMethodinit

Initialize parts of decoder that are dependent on dimension of encoder tokens. Should be called when setting up MultiMAE. :p

src/croco/models/dpt_block.py:380

↓ 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 callersMethodprepare_tokens_with_masks

(self, x, masks=None)

src/slamformer/models/dinov2/models/vision_transformer.py:215

↓ 4 callersFunctionreverse_imagenet_normalize

Reverses ImageNet normalization in an input image.

slam/rerun_helper/generic_utils.py:145

↓ 4 callersMethodstate_dict

(self)

src/croco/utils/misc.py:316

↓ 3 callersMethod__init__

Init. Args: scale_factor (float): scaling mode (str): interpolation mode

src/croco/models/dpt_block.py:250

↓ 3 callersFunction_crop

(img, sy, sx)

src/croco/stereoflow/engine.py:360

↓ 3 callersFunction_read_png_disp

(filename, coef=1.0)

src/croco/stereoflow/datasets_stereo.py:838

↓ 3 callersFunction_run

(command)

src/slamformer/models/dinov2/utils/utils.py:49

↓ 3 callersMethodclose

(self)

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

↓ 3 callersFunctioncompute_camera_pose_opencv_convention

(camera_position, camera_orientation)

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:26

↓ 3 callersMethodforward

(self, x: Tensor, attn_bias=None)

src/slamformer/models/layers/attention.py:58

↓ 3 callersFunctionget_cluster_type

(cluster_type: Optional[ClusterType] = None)

src/slamformer/models/dinov2/utils/cluster.py:31

↓ 3 callersMethodload_model

(self, ckpt=None, ckpt_raw=None)

slam/demo.py:137

↓ 3 callersMethodlog_every

( self, iterable, print_freq, accelerator: Accelerator, header=None, max_iter=None, start_step=0,

src/croco/utils/misc.py:131

↓ 2 callersMethod__call__

(self, img1, img2)

src/croco/datasets/transforms.py:13

↓ 2 callersMethod__init__

( self, dim: int, num_heads: int, mlp_ratio: float = 4.0, qkv_bias: bo

src/slamformer/models/layers/block.py:41

↓ 2 callersMethod__init__

( self, dim: int, num_heads: int = 8, qkv_bias: bool = False, proj_bia

src/slamformer/models/layers/attention.py:39

↓ 2 callersMethod__init__

(self, eps: float = 1e-8)

slam/rerun_helper/geometry_utils.py:65

↓ 2 callersMethod_add_point_cloud

Add a point cloud to the scene.

slam/visualize_results.py:261

↓ 2 callersFunction_computeColor

flow_utils._computeColor(flow): compute color codes for the flow field flow ---- PARAMETERS ---- flow: np.array of dimension (height

src/croco/stereoflow/datasets_flow.py:851

↓ 2 callersMethod_decoder

return_all_blocks: if True, return the features at the end of every block instead of just the features from the la

src/croco/models/croco.py:247

↓ 2 callersMethod_encode_image

image has B x 3 x img_size x img_size do_mask: whether to perform masking or not return_all_blocks: if True, return the featu

src/croco/models/croco.py:210

↓ 2 callersMethod_error

(self, gt, predictions)

src/croco/stereoflow/criterion.py:30

↓ 2 callersFunction_make_dinov2_model_name

(arch_name: str, patch_size: int, num_register_tokens: int = 0)

src/slamformer/models/dinov2/hub/utils.py:17

↓ 2 callersFunction_overlapping

(total, window, overlap=0.5)

src/croco/stereoflow/engine.py:353

↓ 2 callersFunction_read_pfm

(file)

src/croco/stereoflow/datasets_stereo.py:886

↓ 2 callersMethod_resize_flow

(self, flow, scale_x, scale_y, factor=1.0)

src/croco/stereoflow/augmentor.py:290

↓ 2 callersFunction_resize_img

(img, new_size)

src/croco/stereoflow/engine.py:225

↓ 2 callersFunction_resize_stereo_or_flow

(data, new_size)

src/croco/stereoflow/engine.py:229

↓ 2 callersMethodadd_meter

(self, name, meter)

src/croco/utils/misc.py:128

↓ 2 callersMethodapply_rope1d

(self, tokens, pos1d, cos, sin)

src/slamformer/models/layers/pos_embed.py:138

↓ 2 callersMethodapply_rope1d

(self, tokens, pos1d, cos, sin)

src/croco/models/pos_embed.py:152

↓ 2 callersFunctionarea

(sz)

src/croco/datasets/crops/extract_crops_from_images.py:136

↓ 2 callersMethodbackend

(self, final=False)

slam/demo.py:305

↓ 2 callersFunctioncompute_camera_intrinsics

(height, width, hfov)

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:20

↓ 2 callersFunctioncompute_pointcloud_overlaps_scikit

Compute 'overlapping' metrics based on a distance threshold between two point clouds.

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:59

↓ 2 callersFunctiondrop_add_residual_stochastic_depth

( x: Tensor, residual_func: Callable[[Tensor], Tensor], sample_drop_ratio: float = 0.0, )

src/slamformer/models/dinov2/layers/block.py:116

↓ 2 callersFunctiondrop_add_residual_stochastic_depth_list

( x_list: List[Tensor], residual_func: Callable[[Tensor, Any], Tensor], sample_drop_ratio: float =

src/slamformer/models/layers/block.py:186

↓ 2 callersFunctiondrop_add_residual_stochastic_depth_list

( x_list: List[Tensor], residual_func: Callable[[Tensor, Any], Tensor], sample_drop_ratio: float =

src/slamformer/models/dinov2/layers/block.py:187

↓ 2 callersMethodforward

(self, img1, img2)

src/croco/datasets/transforms.py:20

↓ 2 callersMethodfrontendT

(self, img)

src/slamformer/models/slamformer.py:501

↓ 2 callersFunctiongenerate_orientation_noise

(pan_range, tilt_range, roll_range)

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:125

↓ 2 callersFunctionget_1d_sincos_pos_embed_from_grid

embed_dim: output dimension for each position pos: a list of positions to be encoded: size (M,) out: (M, D)

src/slamformer/models/layers/pos_embed.py:52

↓ 2 callersFunctionget_1d_sincos_pos_embed_from_grid

embed_dim: output dimension for each position pos: a list of positions to be encoded: size (M,) out: (M, D)

src/croco/models/pos_embed.py:53

↓ 2 callersFunctionget_2d_sincos_pos_embed

grid_size: int of the grid height and width return: pos_embed: [grid_size*grid_size, embed_dim] or [n_cls_token+grid_size*grid_size, embe

src/croco/models/pos_embed.py:22

↓ 2 callersFunctionget_attn_bias_and_cat

this will perform the index select, cat the tensors, and provide the attn_bias from cache

src/slamformer/models/layers/block.py:162

↓ 2 callersFunctionget_attn_bias_and_cat

this will perform the index select, cat the tensors, and provide the attn_bias from cache

src/slamformer/models/dinov2/layers/block.py:163

↓ 2 callersFunctionis_dist_avail_and_initialized

()

src/croco/utils/misc.py:220

↓ 2 callersMethodis_other_pointcloud_overlapping

Check if a viewpoint is valid and overlaps significantly with a reference one.

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:301

↓ 2 callersMethodjitter_one

( self, img, fn_idx, brightness_factor, contrast_factor, satur

src/croco/datasets/transforms.py:47

↓ 2 callersMethodload_checkpoint

(self)

slam/demo.py:119

↓ 2 callersFunctionload_image

(impath)

src/croco/datasets/pairs_dataset.py:11

↓ 2 callersFunctionlook_at_for_habitat

(eye, center, up, return_cam2world=True)

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:119

↓ 2 callersFunctionmake_2tuple

(x)

src/slamformer/models/dinov2/layers/patch_embed.py:16

↓ 2 callersMethodmap_add

(self, token_kf)

slam/demo.py:365

↓ 2 callersFunctionnormalized_view_plane_uv

UV with left-top corner as (-width / diagonal, -height / diagonal) and right-bottom corner as (width / diagonal, height / diagonal)

src/slamformer/models/layers/conv_head.py:11

↓ 2 callersFunctionprepare_crop

(img, rect, rot=0)

src/croco/datasets/crops/extract_crops_from_images.py:141

↓ 2 callersFunctionrotate_target_dim_to_last_axis

(x, target_dim=3)

src/slamformer/utils/basic.py:133

↓ 2 callersMethodsample_random_viewpoint

Sample a random viewpoint using the navmesh

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:259

↓ 2 callersFunctionsave_on_master

(accelerator: Accelerator, *args, **kwargs)

src/croco/utils/misc.py:244

↓ 2 callersFunctionsetup

Create configs and perform basic setups.

src/slamformer/models/dinov2/utils/config.py:63

↓ 2 callersMethodsetup

(self, croconet)

src/croco/models/head_downstream.py:44

↓ 2 callersFunctionsetup_for_distributed

This function disables printing when not in master process

src/croco/utils/misc.py:203

↓ 2 callersFunctionsplit_prediction_conf

(predictions, with_conf=False)

src/croco/stereoflow/engine.py:18

↓ 2 callersMethodstep

(self, timestamp, image)

slam/demo.py:355

↓ 2 callersFunctiontiled_pred

( model, criterion, img1, img2, gt, overlap=0.5, bad_crop_thr=0.05, downscale=

src/croco/stereoflow/engine.py:243

↓ 2 callersFunctionto_homogeneous

Converts tensor to homogeneous coordinates by adding ones to the specified dimension

slam/rerun_helper/geometry_utils.py:11

↓ 2 callersMethodupdate

(self, value, n=1)

src/croco/utils/misc.py:43

↓ 2 callersFunctionwriteFlowKitti

(filename, uv)

src/croco/stereoflow/datasets_flow.py:756

↓ 2 callersFunctionwritePFM

(file, image, scale=1)

src/croco/stereoflow/datasets_stereo.py:924

↓ 1 callersMethodDivPrune

Diversity-based token pruning. This is a thin wrapper to reuse the existing incremental diversity selector. Returns selected indices

src/slamformer/models/slamformer.py:216

↓ 1 callersMethodKFT

(self, img)

src/slamformer/models/slamformer.py:476

↓ 1 callersMethodPruner

(self, visual_feature_vectors, image_feature_length, cosine_matrix=None, threshold_ratio=0.1, threshold_terms=

src/slamformer/models/slamformer.py:185

↓ 1 callersMethod__init__

(self, patch_size, dec_embed_dim, output_dim=3,)

src/slamformer/models/layers/transformer_head.py:64

↓ 1 callersMethod__init__

( self, num_features: int, dim_in: int, dim_out: List[int], dim_pro

src/slamformer/models/layers/conv_head.py:63

↓ 1 callersMethod__init__

(self, dim=512)

src/slamformer/models/layers/camera_head.py:33

next →1–100 of 601, ranked by callers