Private function to rotate bounding boxes and points. Args: input_dict (dict): Result dict from loading pipeline. Return

Private function to rotate bounding boxes and points. Args: input_dict (dict): Result dict from loading pipeline. rot

Get the 3D rotation matrix. Args: rot_mat_T (np.ndarray): Transposed rotation matrix. angle (float): Rotation angle. axis

Sample specific number of ground truths and return indices. Args: num (int): Sampled number. Returns: list[i

Randomly sample some negative samples.

Randomly sample some positive samples.

Private function to scale bounding boxes and points. Args: input_dict (dict): Result dict from loading pipeline. Returns

r""" Computes scaled dot product attention on query, key and value tensors, using an optional attention mask if passed, and applying dropout

Train semantic head. Args: x (torch.Tensor): Point-wise semantic features for segmentation voxels_dict (dict): Contai

Set flag according to image aspect ratio. Images with aspect ratio greater than 1 will be set as group 1, otherwise group 0. In 3D da

Sampling points in a sliding window fashion. First sample patches to cover all the input points. Then sample points in each patch to

Get indexes based on scores. Args: scores (torch.Tensor): scores with the shape of [B, N, W, H]. K (int): Number to b

Private function to translate bounding boxes and points. Args: input_dict (dict): Result dict from loading pipeline. Ret

Private function to translate points. Args: input_dict (dict): Result dict from loading pipeline. trans_factor (np.nd

Args: voxel_feat: shape=[N, C] coors: [N, 4] Return: batch_canvas:, shape=[B, C, ny, nx]

Convert the rotation difference to difference in sine function. Args: boxes1 (torch.Tensor): Original Boxes in shape (NxC), where

Pre-compute features with weight matrices in weight bank. This function is used before cuda op assign_score_withk in CUDA version PAConv. Arg

Perform weighted sum to aggregate output features according to scores. This function is used in non-CUDA version of PAConv. Compared to the c

Test with augmentations. Args: points (list[torch.Tensor]): List of points of shape [B, N, 3+C]. img_metas (list[list

Test function with augmentation, image network pretrain. May refer to https://github.com/open- mmlab/mmdetection/blob/master/mmdet/mod

Test function of point cloud branch with augmentaiton. The function implementation process is as follows: - step 1: map features

Test function of point cloud branch with augmentaiton.

Calculate overlap between two set of axis aligned 3D bboxes. If ``is_aligned `` is ``False``, then calculate the overlaps between each bbox of

Calculate the IoU loss (1-IoU) of two set of axis aligned bounding boxes. Note that predictions and targets are one-to-one corresponded. Args

Transform 2d bbox according to img_meta. Args: img_meta(dict): Meta info regarding data transformation. bbox_2d (torch.Tensor): S

compute distance between box corners to replace iou Args: crnr1, crnr2: Nx3 points of box corners in camera axis (y points down)

Calculate 3D IoU using cuda implementation. Note: This function calculates the IoU of 3D boxes based on their volumes. IoU calcul

Transform 3D boxes. Args: boxes (np.ndarray): 3D boxes to be transformed. loc_transform (np.ndarray): Location transform to be ap

Covert boxes in camera coordinate to lidar coordinate. Args: data (np.ndarray, shape=[N, 7]): Boxes in camera coordinate. r_rect

Build data config for loading visualization data.

Build middle level encoder.

Build roi layers using `layer_cfg`

Calculate the Euclidian distance between two sets of points. Args: xyz1 (torch.Tensor): (N, 3), the first set of points. xyz2 (to

Fast iou algorithm. this function can be used independently to do result analysis. Must be used in CAMERA coordinate system. Args: gt

Convert transformation matrix in Cartesian coordinates to homogeneous format. Args: mat (np.ndarray): Transformation matr

Center to minmax. Args: centers (np.ndarray): Center points. dims (np.ndarray): Dimensions. origin (list or array or floa

Whether the mean of points to plane distance is lower than thresh. Args: plane_equ (torch.Tensor): Plane to be checked.

Check whether is a horizontal plane. Args: points (torch.Tensor): Points of input. Returns: Bool: Flag of re

Class agnostic nms for single batch. Args: mlvl_bboxes (torch.Tensor): Bboxes from Multi-level. mlvl_bboxes_for_nms (

Classify each points.

Combine predictions in waymo format for each sample together. Args: pathnames (str): Paths to save predictions. Returns:

Compute loss of primitive module. Args: primitive_center (torch.Tensor): Pridictions of primitive center. primitive_s

Concat data_infos from several datasets to form self.data_infos. Args: data_infos (list[list[dict]])

Concat scene_idxs from several datasets to form self.scene_idxs. Needs to manually add offset to scene_idxs[1, 2, ...]. Args:

Convert keys in pycls pretrained RegNet models to mmdet style.

Convert boxes from `src` mode to `dst` mode. Args: box (tuple | list | np.ndarray | torch.Tensor | BaseInstance3D

Convert self to ``dst`` mode. Args: dst (:obj:`BoxMode`): the target Box mode rt_mat (np.ndarray | torch.Tensor): The

Convert the predicted boxes into valid ones. Args: box_dict (dict): Box dictionaries to be converted. - boxes_3d

Convert the predicted boxes into valid ones. Args: box_dict (dict): Box dictionaries to be converted. - boxes_3d

Convert the boxes into valid format. Args: box_dict (dict): Bounding boxes to be converted. - boxes_3d (:obj:``L

Convert the boxes into valid format. Args: box_dict (dict): Bounding boxes to be converted. - boxes_3d (:obj:``L

Transform 2d pixel coordinates according to img_meta. Args: img_meta(dict): Meta info regarding data transformation. coord_2d (to

convert 3d box corners from corner function above to surfaces that normal vectors all direct to internal. Args: corners (np.ndarray):

Convert 3d box corners from corner function above to surfaces that normal vectors all direct to internal. Args: corners (np.ndarray):

Create folder for data preprocessing.

Create folder for data conversion.

Forward of upsample and residual block. Args: x_lateral (:obj:`SparseConvTensor`): Lateral tensor. x_bottom (:obj:`Sp

Convert depth map to points. Args: depth (np.array, shape=[H, W]): Depth map which the row of [0~`trunc_pixel`] are truncated

Args: det11: (np.ndarray): [xy, yaw[:, None], bot[:, None], height[:, None]] Returns: np.ndarray: Converted boxes in xyzwhlr

Compute rotated iou on device. Args: rbox1 (np.ndarray, shape=[5]): Rotated 2d box. rbox2 (np.ndarray, shape=[5]): Rotated 2d box