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

( self, in_features: int, hidden_features: Optional[int] = None, out_features:

src/slamformer/models/dinov2/layers/swiglu_ffn.py:15

Build network for binocular downstream task It takes an extra argument head, that is called with the features and a dictionary img_i

src/croco/models/croco_downstream.py:78

↓ 1 callersMethod__init__

(self, config: CrocoConfig)

src/croco/models/croco.py:64

↓ 1 callersMethod_add_camera

Add a camera frustum to the scene.

slam/visualize_results.py:297

↓ 1 callersFunction_append_camera_parameters

Add camera parameters to the observation dictionnary produced by Habitat-Sim In-place modifications.

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:80

↓ 1 callersMethod_build_cache

(self)

src/croco/stereoflow/datasets_flow.py:440

↓ 1 callersMethod_build_cache

(self)

src/croco/stereoflow/datasets_stereo.py:287

↓ 1 callersFunction_build_mlp

(nlayers, in_dim, bottleneck_dim, hidden_dim=None, use_bn=False, bias=True)

src/slamformer/models/dinov2/layers/dino_head.py:44

↓ 1 callersMethod_compute_metrics

(self)

src/croco/stereoflow/criterion.py:220

↓ 1 callersMethod_compute_metrics

(self)

src/croco/stereoflow/criterion.py:333

↓ 1 callersMethod_generate_camera_colors

Generate gradient colors for cameras.

slam/visualize_results.py:167

↓ 1 callersMethod_get_intermediate_layers_chunked

(self, x, n=1)

src/slamformer/models/dinov2/models/vision_transformer.py:292

↓ 1 callersMethod_get_intermediate_layers_not_chunked

(self, x, n=1)

src/slamformer/models/dinov2/models/vision_transformer.py:280

↓ 1 callersFunction_get_num_layer_for_vit

(var_name, enc_depth, dec_depth)

src/croco/utils/misc.py:512

↓ 1 callersMethod_get_pad

(self, size)

src/slamformer/models/dinov2/hub/utils.py:28

↓ 1 callersFunction_guess_cluster_type

()

src/slamformer/models/dinov2/utils/cluster.py:18

↓ 1 callersMethod_init_weights

(self, m)

src/croco/models/croco.py:200

↓ 1 callersMethod_lazy_initialization

(self)

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:195

↓ 1 callersFunction_load_model_and_criterion

(model_path, do_load_metrics, device)

src/croco/stereoflow/test.py:71

↓ 1 callersMethod_load_or_build_cache

(self)

src/croco/stereoflow/datasets_flow.py:126

↓ 1 callersMethod_load_or_build_cache

(self)

src/croco/stereoflow/datasets_stereo.py:144

↓ 1 callersMethod_make_output_block

(self, dim_in: int, dim_out: int, dim_times_res_block_hidden: int, last_res_blocks: int, last_conv_channels: i

src/slamformer/models/layers/conv_head.py:106

↓ 1 callersMethod_make_upsampler

(self, in_channels: int, out_channels: int)

src/slamformer/models/layers/conv_head.py:98

↓ 1 callersFunction_ntuple

(n)

src/croco/models/blocks.py:25

↓ 1 callersMethod_prepare_data

(self)

src/croco/stereoflow/datasets_flow.py:401

↓ 1 callersMethod_prepare_data

(self)

src/croco/stereoflow/datasets_stereo.py:273

↓ 1 callersFunction_print_repr_dataset

(d)

src/croco/stereoflow/train.py:299

↓ 1 callersMethod_random_color

(self, img1, img2)

src/croco/stereoflow/augmentor.py:205

↓ 1 callersMethod_random_crop

(self, img1, img2, disp)

src/croco/stereoflow/augmentor.py:122

↓ 1 callersMethod_random_rotate_shift_right

(self, img2)

src/croco/stereoflow/augmentor.py:141

↓ 1 callersMethod_random_scale

(self, img1, img2, disp)

src/croco/stereoflow/augmentor.py:59

↓ 1 callersMethod_random_vflip

(self, img1, img2, disp)

src/croco/stereoflow/augmentor.py:133

↓ 1 callersMethod_regenerate_point_clouds

Regenerate all point clouds with current settings.

slam/visualize_results.py:249

↓ 1 callersFunction_replace

Advanced string replacement. Given a text: - replace all elements in src by the corresponding element in tgt - remove all elements in rm

src/croco/utils/misc.py:481

↓ 1 callersFunction_save_batch

( pred, gt, pairnames, dataset, task, save, outdir, time, submission_dir=None )

src/croco/stereoflow/test.py:103

↓ 1 callersMethod_set_decoder

( self, enc_embed_dim, dec_embed_dim, dec_num_heads, dec_depth,

src/croco/models/croco.py:156

↓ 1 callersMethod_set_mask_generator

(self, num_patches, mask_ratio)

src/croco/models/croco.py:150

↓ 1 callersMethod_set_patch_embed

(self, img_size=224, patch_size=16, enc_embed_dim=768)

src/croco/models/croco.py:147

↓ 1 callersMethod_set_prediction_head

(self, dec_embed_dim, patch_size)

src/croco/models/croco.py:188

↓ 1 callersMethod_setup_gui

Setup interactive GUI elements.

slam/visualize_results.py:176

↓ 1 callersMethodadapt_tokens

(self, encoder_tokens)

src/croco/models/dpt_block.py:471

↓ 1 callersMethodadd_batch

(self, predictions, gt)

src/croco/stereoflow/criterion.py:255

↓ 1 callersFunctionadd_residual

(x, brange, residual, residual_scale_factor, scaling_vector=None)

src/slamformer/models/layers/block.py:147

↓ 1 callersFunctionadd_residual

(x, brange, residual, residual_scale_factor, scaling_vector=None)

src/slamformer/models/dinov2/layers/block.py:148

↓ 1 callersMethodanimate

Setup animation and playback controls.

slam/visualize_results.py:331

↓ 1 callersFunctionapply_scaling_rules_to_cfg

(cfg)

src/slamformer/models/dinov2/utils/config.py:21

↓ 1 callersFunctionarg_parser

()

src/croco/datasets/crops/extract_crops_from_images.py:17

↓ 1 callersFunctionarray_to_pil

Converts a NumPy array to a PIL image. Automatically: - Squeezes dimensions of size 1. - Moves the channel dimension (if it has s

src/slamformer/utils/basic.py:105

↓ 1 callersMethodbackendT

(self, hidden_F)

src/slamformer/models/slamformer.py:527

↓ 1 callersMethodbackward

(ctx, grad_res)

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

↓ 1 callersFunctionbuild_model

(args, only_teacher=False, img_size=224)

src/slamformer/models/dinov2/models/__init__.py:14

↓ 1 callersMethodcheckpoint_uses_conv_head

(self, ckpt)

slam/demo.py:134

↓ 1 callersMethodcolor_transform

Photometric augmentation

src/croco/stereoflow/augmentor.py:272

↓ 1 callersFunctioncompute_pointmap

Compute a HxWx3 pointmap in camera frame from a HxW depth map.

src/croco/datasets/habitat_sim/multiview_habitat_sim_generator.py:32

↓ 1 callersMethodconvert_pose_to_4x4

(self, B, out_r, out_t, device)

src/slamformer/models/layers/camera_head.py:66

↓ 1 callersFunctioncreate_commandline

Create a commandline string to generate a scene.

src/croco/datasets/habitat_sim/generate_multiview_images.py:160

↓ 1 callersFunctioncroco_args_from_ckpt

(ckpt)

src/croco/models/croco_downstream.py:13

↓ 1 callersFunctioncrop_generator

()

src/croco/stereoflow/engine.py:288

↓ 1 callersFunctioncrop_image_to_target_ratio

Crops an image to satisfy a target aspect ratio.

slam/rerun_helper/generic_utils.py:202

↓ 1 callersFunctiondefault_setup

(args)

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

↓ 1 callersFunctiondepthmap_to_camera_coordinates

Args: - depthmap (HxW array): - camera_intrinsics: a 3x3 matrix Returns: pointmap of absolute coordinates (HxWx3 arra

src/slamformer/utils/geometry.py:80

↓ 1 callersFunctiondinov2_vitl14_reg

DINOv2 ViT-L/14 model with registers (optionally) pretrained on the LVD-142M dataset.

src/slamformer/models/dinov2/hub/backbones.py:128

↓ 1 callersFunctiondisp_to_tensor

(disp)

src/croco/stereoflow/datasets_stereo.py:51

↓ 1 callersFunctiondnames_to_image_pairs

dnames: list of datasets with image pairs, separated by +

src/croco/datasets/pairs_dataset.py:90

↓ 1 callersFunctiondrop_path

(x, drop_prob: float = 0.0, training: bool = False)

src/slamformer/models/dinov2/layers/drop_path.py:14

↓ 1 callersFunctiondrop_path

Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).

src/croco/models/blocks.py:37

↓ 1 callersMethodencode_image_pairs

run encoder for a pair of images it is actually ~5% faster to concatenate the images along the batch dimension than to encode them se

src/croco/models/croco_downstream.py:101

↓ 1 callersMethodfinalize_submission

(self, outdir)

src/croco/stereoflow/datasets_flow.py:522

↓ 1 callersFunctionflowMaxNorm

flow_utils.flowMaxNorm(flow): return the maximum of the l2-norm of the given flow ---- PARAMETERS ---- flow: the flow ---- OUTPU

src/croco/stereoflow/datasets_flow.py:839

↓ 1 callersFunctionflowToColor

flow_utils.flowToColor(flow): return a color code flow field, normalized based on the maximum l2-norm of the flow flow_utils.flowToColor(flow

src/croco/stereoflow/datasets_flow.py:808

↓ 1 callersFunctionflow_to_tensor

(disp)

src/croco/stereoflow/datasets_flow.py:36

↓ 1 callersMethodforward

(self, x: Tensor)

src/slamformer/models/layers/block.py:87

↓ 1 callersMethodforward

(self, x: Tensor)

src/slamformer/models/dinov2/layers/block.py:88

↓ 1 callersMethodforward

(self, x: Tensor, attn_bias=None)

src/slamformer/models/dinov2/layers/attention.py:56

↓ 1 callersMethodforward_features

(self, x, masks=None)

src/slamformer/models/dinov2/models/vision_transformer.py:259

↓ 1 callersMethodforward_features_list

(self, x_list, masks_list)

src/slamformer/models/dinov2/models/vision_transformer.py:236

↓ 1 callersMethodforward_nested

x_list contains a list of tensors to nest together and run

src/slamformer/models/layers/block.py:210

↓ 1 callersMethodforward_nested

x_list contains a list of tensors to nest together and run

src/slamformer/models/dinov2/layers/block.py:211

↓ 1 callersMethodfrontend

(self, image)

slam/demo.py:202

↓ 1 callersFunctiongenerate_multiview_images_for_scene

Generate tuples of overlapping views for a given scene. generate_depth: generate depth images and camera parameters.

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

↓ 1 callersFunctiongenerate_multiview_images_from_metadata

Generate images from a metadata file for reproducibility purposes.

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

↓ 1 callersFunctionget_2d_sincos_pos_embed_from_grid

(embed_dim, grid)

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

↓ 1 callersFunctionget_2d_sincos_pos_embed_from_grid

(embed_dim, grid)

src/croco/models/pos_embed.py:42

↓ 1 callersFunctionget_args_parser

()

src/croco/pretrain.py:37

↓ 1 callersFunctionget_args_parser

()

src/croco/stereoflow/train.py:40

↓ 1 callersFunctionget_args_parser

()

src/croco/stereoflow/test.py:31

↓ 1 callersFunctionget_branges_scales

(x, sample_drop_ratio=0.0)

src/slamformer/models/layers/block.py:139

↓ 1 callersFunctionget_branges_scales

(x, sample_drop_ratio=0.0)

src/slamformer/models/dinov2/layers/block.py:140

↓ 1 callersFunctionget_cfg_from_args

(args)

src/slamformer/models/dinov2/utils/config.py:40

↓ 1 callersFunctionget_checkpoint_path

(cluster_type: Optional[ClusterType] = None)

src/slamformer/models/dinov2/utils/cluster.py:38

↓ 1 callersMethodget_cos_sin

(self, D, seq_len, device, dtype)

src/slamformer/models/layers/pos_embed.py:122

↓ 1 callersMethodget_cos_sin

(self, D, seq_len, device, dtype)

src/croco/models/pos_embed.py:134

↓ 1 callersFunctionget_grad_norm_

(parameters, norm_type: float = 2.0)

src/croco/utils/misc.py:374

↓ 1 callersFunctionget_pair_transforms

(transform_str, totensor=True, normalize=True)

src/croco/datasets/transforms.py:101

↓ 1 callersFunctionget_parser

()

slam/demo.py:508

↓ 1 callersFunctionget_path

(idx)

src/croco/datasets/crops/extract_crops_from_images.py:98

↓ 1 callersFunctionget_pixel

(H, W)

src/slamformer/utils/geometry.py:42

↓ 1 callersMethodget_results

(self)

src/croco/stereoflow/criterion.py:349

↓ 1 callersFunctionget_slurm_partition

(cluster_type: Optional[ClusterType] = None)

src/slamformer/models/dinov2/utils/cluster.py:61

↓ 1 callersFunctionget_vit_lr_decay_rate

Calculate lr decay rate for different ViT blocks. Args: name (string): parameter name. lr_decay_rate (float): base lr decay r

src/slamformer/models/dinov2/utils/param_groups.py:13

← previous next →101–200 of 601, ranked by callers