Compute regression and classification targets for anchors. Args: anchor_list (list[list]): Multi level anchors of each image.

Test function with augmentation, stage 2. Args: points (list[list[torch.Tensor]], optional): the outer list i

Convert 3D detection results to kitti format for evaluation and test submission. Args: net_outputs (list[np.ndarray])

Convert 3D detection results to kitti format for evaluation and test submission. Args: net_outputs (list[np.ndarray])

Convert results to kitti format for evaluation and test submission. Args: net_outputs (List[np.ndarray]): list of array storin

Convert 2D detection results to kitti format for evaluation and test submission. Args: net_outputs (list[np.ndarray])

Convert 2D detection results to kitti format for evaluation and test submission. Args: net_outputs (list[np.ndarray])

Box collision test. Args: boxes (np.ndarray): Corners of current boxes. qboxes (np.ndarray): Boxes to be avoid colliding.

Convert the box from camera to global coordinate. Args: info (dict): Info for a specific sample data, including the calib

Convert kitti locations, dimensions and angles to corners. format: center(xy), dims(xy), angles(counterclockwise when positive) Args:

Convert kitti locations, dimensions and angles to corners. Args: centers (np.ndarray): Locations in kitti label file with shape (N, 3)

Collect the information of the running environments.

This function would modify some filed to keep the compatibility of config. For example, it will move some args which will be deprecated to

1x1 convolution with padding

3x3 convolution with padding

Convert action.

Convert boxes from `src` mode to `dst` mode. Args: box (tuple | list | np.ndarray | torch.Tensor | :obj:`Base

Convert range images to point cloud. Args: frame (:obj:`Frame`): Open dataset frame. range_images (dict): Mapping

Generate relative box corners based on length per dim and origin point. Args: dims (np.ndarray, shape=[N, ndim]): Array of length per

Given the raw data, generate the ground truth database. Args: dataset_class_name (str): Name of the input dataset. data_path

Decode center, size residuals and class to corners. Only useful for axis-aligned bounding boxes, so angle isn't considered. Args:

str: The device of the boxes are on.

str: The device of the points are on.

Export original files to vert, ins_label, sem_label and bbox file. Args: mesh_file (str): Path of the mesh_file. agg_file (st

Extract image augmentation information from img_meta. Args: img_meta(dict): Meta info regarding data transformation. tensor(t

Extract features from points.

Extract features from points.

Extract features from points. Args: points (list[Tensor]): Raw point clouds. Returns: SparseTensor: Vo

Extract features from points.

Extract features from images and points.

Extract 3d features from the backbone -> fpn -> 3d projection. -> 3d neck -> bbox_head. Args: img (torch.Tensor): I

Directly extract features from the backbone+neck. Args: points (torch.Tensor): Input points. Returns:

Directly extract features from the backbone+neck. Args: points (torch.Tensor): Input points.

Flip the boxes in BEV along given BEV direction. Args: bev_direction (str, optional): Direction by which to flip.

Format the results to pkl file. Args: outputs (list[dict]): Testing results of the dataset. pklfile_prefix (str):

Format the results to json (standard format for COCO evaluation). Args: results (list[dict]): Testing results of the dataset.

Placeholder of forward function.

Calls either simple_test or aug_test depending on the length of outer list of points. If len(points) == 1, call simple_test. Otherwise

Get the 2D annotation records for a given info. Args: info: Information of the given sample data. occluded: Integer (0, 1, 2,

Get annotation info according to the given index. Args: index (int): Index of the annotation data to get. Returns:

Get annotation info according to the given index. Args: index (int): Index of the annotation data to get. Returns:

Get annotation info according to the given index. Args: index (int): Index of the annotation data to get. Returns:

Get available scenes from the input nuscenes class. Given the raw data, get the information of available scenes for further info generatio

Transform network outputs for a batch into bbox predictions. Args: all_cls_scores (Tensor): Classification score of all

Get class names of current dataset. Args: classes (Sequence[str] | str): If classes is None, use default CLAS

Get data info according to the given index. Args: index (int): Index of the sample data to get. Returns:

Encode direction to 0 ~ num_bins-1. Args: anchors (torch.Tensor): Concatenated multi-level anchor. reg_targets (torch.Tensor)

Encode direction to 0 ~ num_bins-1. Args: reg_targets (torch.Tensor): Bbox regression targets. dir_offset (int, o

Calculate the locations of the frustum points in the lidar coordinate system. Args: rots (torch.Tensor): Rotation fro

Calculate loss of voting module. Args: seed_points (torch.Tensor): Coordinate of the seed points. vote_points (to

Create boolean mask by actually number of a padded tensor. Args: actual_num (torch.Tensor): Actual number of points in each voxel.

Decode box predictions and get projected 2D attributes. Args: bbox_preds (list[Tensor]): Box predictions for each scale

Obtain image features using points. Args: img_meta (dict): Meta info. coord_type (str): 'DEPTH' or 'CAMERA' or 'LIDAR'.

Get the valid radios of feature maps of all level.

Check whether the points are in the given range. Args: point_range (list | torch.Tensor): The range of point

Inference with slide/whole style. Args: points (torch.Tensor): Input points of shape [B, N, 3+C]. img_metas (list

Initialize weights of shared MLP layers and BN layers.

Load annotations from ann_file. Args: ann_file (str): Path of the annotation file. Returns: list[dict]

Load annotations from ann_file. Args: ann_file (str): Path of the annotation file. Returns: list[dict]

Loss function. Args: all_cls_scores (Tensor): Classification score of all decoder layers, has shape

Prepare data related to Lyft dataset. Related data consists of '.pkl' files recording basic infos. Although the ground truth database and

Map voxel features to its corresponding points. Args: pts_coors (torch.Tensor): Voxel coordinate of each point. v

Map the centers of voxels to its corresponding points. Args: pts_coors (torch.Tensor): The coordinates of each points, shape

Prepare data related to nuScenes dataset. Related data consists of '.pkl' files recording basic infos, 2D annotations and groundtruth data

Convert the output to the box class in the nuScenes. Args: detection (dict): Detection results. - boxes_3d (:obj:`BaseI

Calculate 3D overlaps of two boxes. Note: This function calculates the overlaps between ``boxes1`` and ``boxes2``

Obtain image features using points. Args: img_meta (dict): Meta info. img_features (torch.Tensor): 1 x C x H x W image featur

Check points is in 3d convex polygons. Args: points (np.ndarray): Input points with shape of (num_points, 3). polygon_surface

Get the intersection of the convex hull of the reprojected bbox corners and the image canvas, return None if no intersection. Args:

Flip 3D data randomly. Args: input_dict (dict): Result dict from loading pipeline. direction (str, optional): Fli

Rotate points with the given rotation matrix or angle. Args: rotation (float, np.ndarray, torch.Tensor): Rotation matrix

Rotated box iou running in gpu. 500x faster than cpu version (take 5ms in one example with numba.cuda code). convert from [this project]( ht