Get gaussian masked heatmap. Args: heatmap (torch.Tensor): Heatmap to be masked. center (torch.Tensor): Center coord of the heatm

Drop irrelevant ground truths by name. Args: gt_names (list[str]): Names of ground truths. used_classes (list[str]):

To make it clear and easy to follow, we do not use loop to process two shifts.

Generic functions to compute precision/recall for object detection for a single class. Args: pred (dict): Predictions mapping from im

Evaluate mAP and recall. Generic functions to compute precision/recall for object detection for multiple classes. Args: pred

Evaluation in KITTI protocol. Args: results (list[dict]): Testing results of the dataset. metric (str | list[str]): M

Evaluation in KITTI protocol. Args: results (list[dict]): Testing results of the dataset. metric (str | list[str]): M

Evaluate. Evaluation in indoor protocol. Args: results (list[dict]): List of results. metric (str | list[str

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

Directly extract features from the backbone+neck.

Extract point and image features of multiple samples.

Extract features of images.

Extract features from multiple images. Args: imgs (list[torch.Tensor]): A list of images. The images are augmente

Extract features of points.

Extract features of points from multiple samples.

Compute the confusion matrix for every batch. Args: preds (np.ndarray): Prediction labels of points with shape of (num_points, )

Apply dynamic voxelization to points. Args: points (list[torch.Tensor]): Points of each sample. Returns: tor

Apply dynamic voxelization to points. Args: points (list[torch.Tensor]): Points of each sample. Returns: tor

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

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

r"""Format the results to txt file. Refer to `ScanNet documentation <http://kaldir.vc.in.tum.de/scannet_benchmark/documentation>`_. A

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

forward. Args: xyz (Tensor): (B, N, 3) xyz coordinates of the features. new_xyz (Tensor): Ignored. featur

RoIAwarePool3d module forward. Args: rois (torch.Tensor): [N, 7],in LiDAR coordinate, (x, y, z) is the bottom cen

Forward function. Args: x (torch.Tensor): of shape (N, C, N_x, N_y, N_z). Returns: torch.Tensor: 5d feature

Scatter features of single sample. Args: voxel_features (torch.Tensor): Voxel features in shape (N, M, C). coors (tor

Forward function for image branch. This function works similar to the forward function of Faster R-CNN. Args: x (list[to

Forward function for point cloud branch. Args: pts_feats (list[torch.Tensor]): Features of point cloud branch gt_bbox

Scatter features of single sample. Args: voxel_features (torch.Tensor): Voxel features in shape (N, M, C). coors (tor

Args: points (list[torch.Tensor]): the outer list indicates test-time augmentations and inner torch.Tensor should

Forward training function. Args: points (list[torch.Tensor], optional): Points of each sample. Defaults to None.

Freeze all image branch parameters.

Fuse image and point features with mask. Args: features (torch.Tensor): Features of voxel, usually it is the valu

Generate gaussian map. Args: shape (list[int]): Shape of the map. sigma (float): Sigma to generate gaussian map. Defa

Generate one 2D annotation record given various informations on top of the 2D bounding box coordinates. Args: ann_rec (dict): Origina

Generate one 2D annotation record given various informations on top of the 2D bounding box coordinates. Args: ann_rec (dict): Origina

Generate one 2D annotation record given various information on top of the 2D bounding box coordinates. Args: ann_rec (dict): Original

Get the 2D annotation records for a given `sample_data_token`. Args: sample_data_token (str): Sample data token belonging to a camera \

Get the 2D annotation records for a given `sample_data_token`. Args: sample_data_token (str): Sample data token belonging to a camera

Compute the overall accuracy. Args: hist(np.ndarray): Overall confusion martix (num_classes, num_classes ). Returns:

Compute the class average accuracy. Args: hist(np.ndarray): Overall confusion martix (num_classes, num_classes ). 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 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 attribute from predicted index. This is a workaround to predict attribute when the predicted velocity is not reliable. We map the

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

Transform network output for a batch into bbox predictions.

Get bboxes of single branch. Args: cls_scores (torch.Tensor): Class score in single batch. bbox_preds (torch.Tensor):

Get bboxes of single branch. Args: cls_scores (torch.Tensor): Class score in single batch. bbox_preds (torch.Tensor):

Get category distribution of single scene. Args: idx (int): Index of the data_info. Returns: dict[list]: for

Args: points: (N, 3 + (f)) [bxyz + (f)] voxel_idxs: (N, 4) [bxyz] num_points_in_voxel: (N) Return

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

Get class names of current dataset. This function is taken from MMSegmentation. Args: classes (Sequence[str] | str | Non

Returns an array with an average precision per class. Note: Ground truth and predictions should have the following format. .. code-block::

Calculate corner loss of given boxes. Args: pred_bbox3d (torch.FloatTensor): Predicted boxes in shape (N, 7). gt_bbox

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

Get file names of waymo raw data.

Args: batch_win_inds: shape=[N, ]. Indicates which window a voxel belongs to. Window inds is unique is the whole batch. voxel_dro

Args: batch_win_inds: shape=[N, ]. Indicates which window a voxel belongs to. Window inds is unique is the whole batch.

Args: batch_win_inds: shape=[N, ]. Indicates which window a voxel belongs to. Window inds is unique is the whole batch.

Args: batch_win_inds: shape=[N, ]. Indicates which window a voxel belongs to. Window inds is unique is the whole batch.