MCPcopy Create free account

hub / github.com/dcharatan/flowmap / functions

Functions274 in github.com/dcharatan/flowmap

⨍ Functions 274◇ Types & classes 102↳ Endpoints 4
↓ 16 callersMethodcat
(manipulables: list[T], dim: int)
flowmap/misc/manipulable.py:95
↓ 12 callersMethodto
Return a shallow copy of this instance in which all tensors have been moved to the specified device.
flowmap/misc/manipulable.py:25
↓ 11 callersFunctionsample_image_grid
Get normalized (range 0 to 1) coordinates and integer indices for an image.
flowmap/model/projection.py:93
↓ 10 callersFunctionhomogenize_points
Convert batched points (xyz) to (xyz1).
flowmap/model/projection.py:11
↓ 7 callersFunctiondraw_points
( image: Float[Tensor, "3 height width"] | Float[Tensor, "4 height width"], points: Vector, color:
flowmap/visualization/drawing/points.py:13
↓ 7 callersFunctionsave_image
Save an image. Assumed to be in range 0-1.
flowmap/misc/image_io.py:56
↓ 5 callersFunctionadd_label
( image: Float[Tensor, "3 width height"], label: str, font: Path = Path("assets/Inter-Regular.otf"
flowmap/visualization/layout.py:262
↓ 5 callersMethodforward
( self, batch: Batch, flows: Flows, tracks: list[Tracks] | None, model
flowmap/loss/loss.py:31
↓ 5 callersFunctionmake_latex_table
( results: TableRows, metrics: list[str], precisions: list[int], rank_orders: list[Literal[-1,
paper/table.py:64
↓ 5 callersMethodsample
( self, num_frames_in_video: int, device: torch.device, )
flowmap/frame_sampler/frame_sampler.py:24
↓ 5 callersFunctionsanitize_vector
( vector: Vector, dim: int, device: torch.device, )
flowmap/visualization/drawing/types.py:18
↓ 5 callersFunctionunproject
Unproject 2D camera coordinates with the given Z values.
flowmap/model/projection.py:76
↓ 4 callersFunctioncenter_crop_intrinsics
Modify the given intrinsics to account for center cropping.
flowmap/misc/cropping.py:53
↓ 4 callersFunctionhcat
Shorthand for a horizontal linear concatenation.
flowmap/visualization/layout.py:158
↓ 4 callersFunctionload_metrics
( scenes: Iterable[Scene], methods: Iterable[Method], metrics: Iterable[Metric], )
paper/common.py:170
↓ 4 callersFunctionprep_image
(image: FloatImage)
flowmap/misc/image_io.py:37
↓ 4 callersFunctionproject_camera_space
( points: Float[Tensor, "*#batch dim"], intrinsics: Float[Tensor, "*#batch dim dim"], epsilon: flo
flowmap/model/projection.py:49
↓ 4 callersFunctionread_colmap_model
( path: Path, device: torch.device = torch.device("cpu"), reorder: bool = True, )
flowmap/export/colmap.py:114
↓ 3 callersFunctioncompute_backward_flow
Return the positions of all surface points with backward optical flow applied.
flowmap/model/projection.py:165
↓ 3 callersMethodforward
( self, batch: Batch, flows: Flows, global_step: int, )
flowmap/model/model.py:54
↓ 3 callersFunctionget_dataset
( dataset_cfgs: list[DatasetCfg], stage: Stage, frame_sampler_cfg: FrameSamplerCfg,
flowmap/dataset/__init__.py:28
↓ 3 callersFunctionget_flow_predictor
(cfg: FlowPredictorCfg)
flowmap/flow/__init__.py:17
↓ 3 callersFunctionget_image_shape
( original_shape: tuple[int, int], cfg: CroppingCfg, )
flowmap/misc/cropping.py:82
↓ 3 callersFunctionget_typed_config
( data_class: Type[T], cfg: DictConfig, extra_type_hooks: dict = {}, )
flowmap/config/tools.py:16
↓ 3 callersFunctionget_typed_root_config
(cfg_dict: DictConfig, cfg_type: Type[T])
flowmap/config/common.py:54
↓ 3 callersFunctionreproject_points
Transform the input points using the provided relative transformations, then project them using the provided intrinsics. After transformation, the
flowmap/model/projection.py:116
↓ 3 callersFunctionresize_metadata
(path: Path)
scripts/colmap/run_both_at_flowmap_resolution.py:58
↓ 3 callersFunctionrow_gap
(height: str)
paper/generate_splat_comparison.py:124
↓ 3 callersFunctionsanitize_scalar
(scalar: Scalar, device: torch.device)
flowmap/visualization/drawing/types.py:44
↓ 3 callersFunctionsave_svg
(fig: svg.SVG, path: Path)
paper/svg_tools.py:26
↓ 3 callersFunctionseparate_multiple_defaults
Return a function that will pull individual configurations out of a merged dict. For example, the merged dict might look like this: {
flowmap/config/tools.py:28
↓ 3 callersFunctionsplit_videos
( videos: Float[Tensor, "batch frame 3 height width"], )
flowmap/flow/common.py:6
↓ 3 callersMethodto
(self, device: torch.device)
flowmap/model/model_wrapper_overfit.py:44
↓ 2 callersFunction_compute_offset
(base: int, overlay: int, align: Alignment)
flowmap/visualization/layout.py:75
↓ 2 callersFunction_get_cross_dim
(main_axis: Axis)
flowmap/visualization/layout.py:68
↓ 2 callersFunction_get_main_dim
(main_axis: Axis)
flowmap/visualization/layout.py:61
↓ 2 callersFunction_sanitize_color
(color: Color)
flowmap/visualization/layout.py:39
↓ 2 callersFunctionadd_border
( image: Float[Tensor, "channel height width"], border: int = 8, color: Color = 1, )
flowmap/visualization/layout.py:202
↓ 2 callersFunctionalign_surfaces
( surfaces: Float[Tensor, "batch frame height width 3"], backward_flows: Float[Tensor, "batch frame-1
flowmap/model/projection.py:213
↓ 2 callersFunctionapply_color_map_to_image
( image: Float[Tensor, "*batch height width"], color_map: str = "inferno", )
flowmap/visualization/color.py:21
↓ 2 callersFunctioncat
Arrange images in a line. The interface resembles a CSS div with flexbox.
flowmap/visualization/layout.py:118
↓ 2 callersFunctioncenter_crop_images
( images: Float[Tensor, "*batch channel height width"], new_shape: tuple[int, int], )
flowmap/misc/cropping.py:42
↓ 2 callersFunctioncolor_map_depth
( depth: Float[Tensor, "batch height width"], cmap: str = "inferno", invert: bool = True, )
flowmap/visualization/depth.py:7
↓ 2 callersFunctioncompute_ate
( gt: Float[Tensor, "point 3"], predicted: Float[Tensor, "point 3"], )
flowmap/misc/ate.py:7
↓ 2 callersMethodcompute_bidirectional_flow
( self, batch: Batch, flow_shape: tuple[int, int], )
flowmap/flow/flow_predictor.py:82
↓ 2 callersMethodcompute_consistency_mask
( videos: Float[Tensor, "batch frame 3 height width"], flow: Float[Tensor, "batch frame-1 heig
flowmap/flow/flow_predictor.py:60
↓ 2 callersFunctioncompute_patch_cropped_shape
( shape: tuple[int, int], patch_size: int, )
flowmap/misc/cropping.py:30
↓ 2 callersFunctionconvert_to_buffer
(module: nn.Module, persistent: bool = True)
flowmap/misc/nn_module_tools.py:4
↓ 2 callersFunctioncrop_and_resize_batch_for_flow
(batch: Batch, cfg: CroppingCfg)
flowmap/misc/cropping.py:111
↓ 2 callersFunctioncrop_and_resize_batch_for_model
( batch: Batch, cfg: CroppingCfg, )
flowmap/misc/cropping.py:96
↓ 2 callersFunctiondraw_lines
( image: Float[Tensor, "3 height width"] | Float[Tensor, "4 height width"], start: Vector, end: Ve
flowmap/visualization/drawing/lines.py:13
↓ 2 callersFunctionfix_aspect_ratio
When computing losses on normalized image coordinates (width in range [0, 1] and height in range [0, 1]), distances are skewed based on the aspect
flowmap/loss/mapping/mapping.py:9
↓ 2 callersFunctionflow_with_key
( flow: Float[Tensor, "frame height width 2"], )
flowmap/visualization/visualizer_summary.py:21
↓ 2 callersFunctionfocal_lengths_to_intrinsics
( focal_lengths: Float[Tensor, " *batch"], image_shape: tuple[int, int], )
flowmap/model/intrinsics/common.py:6
↓ 2 callersMethodforward
( self, videos: Float[Tensor, "batch frame 3 height width"], )
flowmap/flow/flow_predictor.py:33
↓ 2 callersFunctiongenerate_conversions
( shape: tuple[int, int], device: torch.device, x_range: Optional[Pair] = None, y_range: Optio
flowmap/visualization/drawing/coordinate_conversion.py:19
↓ 2 callersFunctiongenerate_sample_grid
( shape: tuple[int, int], device: torch.device, )
flowmap/visualization/drawing/rendering.py:18
↓ 2 callersFunctionget_extrinsics
Convert the inverse relative transformations from ModelOutput to extrinsics. Each inverse relative transformation transforms points from frame {i
flowmap/model/projection.py:187
↓ 2 callersMethodget_generator
(self, loader_cfg: DataLoaderStageCfg)
flowmap/dataset/data_module_pretrain.py:56
↓ 2 callersFunctionget_image
(path: Path)
paper/generate_splat_comparison.py:128
↓ 2 callersFunctionget_losses
(cfgs: list[LossCfg])
flowmap/loss/__init__.py:13
↓ 2 callersFunctionget_mapping
(cfg: MappingCfg)
flowmap/loss/mapping/__init__.py:15
↓ 2 callersMethodget_persistent
(self, loader_cfg: DataLoaderStageCfg)
flowmap/dataset/data_module_pretrain.py:53
↓ 2 callersFunctionget_visualizers
(cfgs: list[VisualizerCfg])
flowmap/visualization/__init__.py:13
↓ 2 callersMethodload_frame_annotations
(self, category: str)
flowmap/dataset/dataset_co3d.py:150
↓ 2 callersFunctionload_image
( path: Union[Path, str], )
flowmap/misc/image_io.py:70
↓ 2 callersFunctionload_trajectory
(method: Method, scene: Scene)
paper/common.py:164
↓ 2 callersMethodlog_image
( self, key: str, images: list[Any], step: Optional[int] = None, **kwa
flowmap/misc/local_logger.py:35
↓ 2 callersMethodlog_metrics
(self, metrics, step)
flowmap/misc/local_logger.py:31
↓ 2 callersFunctionpatch_crop_batch
(batch: Batch, patch_size: int)
flowmap/misc/cropping.py:71
↓ 2 callersMethodpreprocess_batch
(self, batch_dict: dict)
flowmap/model/model_wrapper_pretrain.py:47
↓ 2 callersFunctionrender_over_image
( image: Float[Tensor, "3 height width"] | Float[Tensor, "4 height width"], color_function: ColorFunct
flowmap/visualization/drawing/rendering.py:136
↓ 2 callersMethodrescale_flow
( flow: Float[Tensor, "batch frame height width 2"], shape: tuple[int, int], )
flowmap/flow/flow_predictor.py:40
↓ 2 callersMethodrescale_mask
( mask: Float[Tensor, "batch frame height width"], shape: tuple[int, int], )
flowmap/flow/flow_predictor.py:50
↓ 2 callersFunctionresize_batch
(batch: Batch, shape: tuple[int, int])
flowmap/misc/cropping.py:19
↓ 2 callersFunctionresize_to_cover
( image: Image.Image, shape: tuple[int, int], )
flowmap/misc/cropping.py:125
↓ 2 callersFunctionresize_to_cover_with_intrinsics
( images: list[Image.Image], shape: tuple[int, int], intrinsics: Float[Tensor, "*batch 3 3"] | Non
flowmap/misc/cropping.py:149
↓ 2 callersFunctionresize_to_resolution
( image: Float[Tensor, "3 height width"], resolution: int, )
flowmap/subsample.py:83
↓ 2 callersFunctionrun_common_training_setup
( cfg: CommonCfg, cfg_dict: DictConfig, )
flowmap/misc/common_training_setup.py:16
↓ 2 callersFunctionrun_dense
(input_path: Path, output_path: Path)
scripts/colmap/run_dense.py:10
↓ 2 callersFunctionrun_sparse
( input_path: Path, output_path: Path, seed: int | None, quality: Quality, data_type: Data
scripts/colmap/run_sparse.py:14
↓ 2 callersFunctionsanitize_pair
(pair: Pair, device: torch.device)
flowmap/visualization/drawing/types.py:61
↓ 2 callersMethodshuffle
(self, lst: list)
flowmap/dataset/dataset_re10k.py:57
↓ 2 callersFunctiontransform_rigid
Apply a rigid-body transformation to points or vectors.
flowmap/model/projection.py:25
↓ 2 callersFunctionvcat
Shorthand for a horizontal linear concatenation.
flowmap/visualization/layout.py:180
↓ 2 callersFunctionversion_to_int
Convert versions of the form vX to X. For example, v12 to 12.
flowmap/misc/wandb_tools.py:8
↓ 2 callersMethodvisualize
( self, batch: Batch, flows: Flows, tracks: list[Tracks] | None, model
flowmap/visualization/visualizer.py:23
↓ 2 callersFunctionwrite_colmap_model
( path: Path, extrinsics: Float[Tensor, "frame 4 4"], intrinsics: Float[Tensor, "frame 3 3"],
flowmap/export/colmap.py:174
↓ 1 callersFunction_intersperse
(iterable: Iterable, delimiter: Any)
flowmap/visualization/layout.py:53
↓ 1 callersFunctionalign_rigid
Compute a rigid transformation that, when applied to p, minimizes the weighted squared distance between transformed points in p and points in q. S
flowmap/model/procrustes.py:7
↓ 1 callersFunctionapply_color_map
( x: Float[Tensor, " *batch"], color_map: str = "inferno", )
flowmap/visualization/color.py:8
↓ 1 callersMethodcat_sliceable
(values: list[Sliceable], dim: int)
flowmap/misc/manipulable.py:109
↓ 1 callersFunctioncolor_function
( xy: Float[Tensor, "point 2"], )
flowmap/visualization/drawing/lines.py:42
↓ 1 callersMethodcompute_correspondence_weights
( self, features_earlier: Float[Tensor, "batch frame channel height width"], features_
flowmap/model/backbone/backbone_midas.py:104
↓ 1 callersFunctioncompute_flows
( batch: Batch, flow_shape: tuple[int, int], device: torch.device, cfg: FlowPredictorCfg, )
flowmap/flow/__init__.py:24
↓ 1 callersFunctioncompute_forward_flow
Return the positions of all surface points with forward optical flow applied.
flowmap/model/projection.py:143
↓ 1 callersFunctioncompute_ranks
( values: Float[np.ndarray, "row col"], orders: Int[np.ndarray, " col"], )
paper/table.py:49
↓ 1 callersFunctioncompute_ranks_for_column
( values: Float[np.ndarray, " row"], order: Literal[-1, 0, 1], )
paper/table.py:24
↓ 1 callersFunctioncompute_track_flow
( surfaces: Float[Tensor, "batch frame height width xyz=3"], extrinsics: Float[Tensor, "batch frame 4
flowmap/model/projection.py:255
↓ 1 callersFunctioncompute_tracks
( batch: Batch, device: torch.device, tracking_cfg: TrackPredictorCfg, precomputation_cfg: Tra
flowmap/tracking/__init__.py:80
next →1–100 of 274, ranked by callers