check if model is in parallel mode.

Args: input_ch(int) : the number of input channel out_features (list[str]): name of the layers whose outputs should

Extract features of points.

Return a point cloud that aggregates multiple sweeps. As every sweep is in a different coordinate frame, we need to map the coordinat

Provides the default colors based on the category names. This method works for the general nuScenes categories, as well as the nuScenes detec

Convert the model into training mode while keep normalization layer freezed.

Save the current checkpoint and delete unwanted checkpoint.

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

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

1x1 convolution with padding

3x3 convolution with padding

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

Input: depths: [B, N, H, W] Output: depths: [B*N*h*w, d]

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

https://github.com/Megvii-BaseDetection/BEVStereo/blob/master/dataset/nusc_mv_det_dataset.py#L48

query_bbox: [B, Q, 10] offset: [B, Q, num_points, 4], normalized by stride

Data preparation for voxel pooling. Args: coor (torch.tensor): Coordinate of points in the lidar space in shape (

Args: model (nn.Module): model to apply EMA. decay (float): ema decay reate. updates (int): counter of EM

Checkpointining without re-entrant autograd Args: function: describes what to run in the forward pass of the model or part of

Get point cloud filename according to the given index. Args: index (int): Index of the point cloud file to get. Returns:

Private function to load point clouds data. Args: pts_filename (str): Filename of point clouds data. Returns:

Private function to load point clouds data. Args: pts_filename (str): Filename of point clouds data. Returns:

Flip bboxes horizontally. Args: bboxes (numpy.ndarray): Bounding boxes, shape (..., 4*k) img_shape (tuple[int]): Imag

r"""Function that computes Binary Focal loss. .. math:: \text{FL}(p_t) = -\alpha_t (1 - p_t)^{\gamma} \, \text{log}(p_t) where:

r"""Checkpoint a model or part of the model Checkpointing works by trading compute for memory. Rather than storing all intermediate activatio

Plot a precision recall curve for the specified class. :param md_list: DetectionMetricDataList instance. :param metrics: DetectionMetrics

Plot the true positive curve for the specified class. :param md_list: DetectionMetricDataList instance. :param metrics: DetectionMetrics

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

Generate the frustum template for each image. Args: depth_cfg (tuple(float)): Config of grid alone depth axis in format

Generate the grid information including the lower bound, interval, and size. Args: x (tuple(float)): Config of grid alone

Decode bboxes. Args: all_cls_scores (Tensor): Outputs from the classification head, \ shape [nb_dec, bs, num_query

Decode bboxes. Args: cls_scores (Tensor): Outputs from the classification head, \ shape [num_query, cls_out_channe

3x3 convolution with padding

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

r"""Criterion that computes Focal loss. According to :cite:`lin2018focal`, the Focal loss is computed as follows: .. math:: \text{FL}(

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

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

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

Input: gt_depths: [B, N, H, W] Output: gt_depths: [B*N*h*w, d]

Returns the data path as well as all annotations related to that sample_data. Note that the boxes are transformed into the current sensor's c

r"""Convert an image from HSV to RGB. The H channel values are assumed to be in the range 0..2pi. S and V are in the range 0..1. Args:

Pre-compute the necessary information in acceleration including the index of points in the final feature. Args: coor (tor