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

Export RGB images to the output path.

Export depth images to the output path.

Export the depth intrinsic matrix to the output path.

Export the intrinsic matrix to the output path.

Export camera poses to the output path.

Extract occupancy annotations of a single dataset to dataset root.

The unified entry for a forward process in both training and test. The method should accept three modes: "tensor", "predict" and "loss":

The unified entry for a forward process in both training and test. The method should accept three modes: "tensor", "predict" and "loss":

The unified entry for a forward process in both training and test. The method should accept three modes: "tensor", "predict" and "loss":

The unified entry for a forward process in both training and test. The method should accept three modes: "tensor", "predict" and "loss":

The unified entry for a forward process in both training and test. The method should accept three modes: "tensor", "predict" and "loss":

Forward with Transformer decoder. The forward procedure of the transformer is defined as: 'pre_transformer' -> 'encoder' -> 'pre_deco

Get frustum corners in camera coordinates. Args: bbox_image (list[int]): box in image coordinates. C (np.ndarray): Intrinsics.

Get the path of the ply file.

Get the root directory of the dataset.

Compute regression and classification targets for a batch image. Outputs from a single decoder layer of a single feature level are used.

Compute targets for final locations for a single scene. Args: points (list[Tensor]): Final locations for all levels.

Compute targets for final locations for a single scene. Args: points (list[Tensor]): Final locations for all levels.

Get voxel-wise segmentation label and point2voxel map. Args: res_coors (Tensor): The voxel coordinates of points, Nx3.

Check whether the points are in the given range. Args: point_range (Tensor or np.ndarray or Sequence[float]): The range of

Indoor Evaluation. Evaluate the result of the detection. Args: gt_annos (list[dict]): Ground truth annotations. dt_annos (li

Initialize a model from config file, which could be a 3D detector or a 3D segmentor. Args: config (str, :obj:`Path`, or :obj:`mmengin

Load a single scene data with multiple RGBD frames.

Calculate losses from a batch of inputs dict and data samples. Args: batch_inputs_dict (dict): The model input dict which include

Calculate losses from a batch of inputs and data samples. Args: batch_inputs_dict (dict): The model input dict which include

Calculate losses from a batch of inputs dict and data samples. Args: batch_inputs_dict (dict): The model input dict which include

Calculate losses from a batch of inputs dict and data samples. Args: batch_inputs_dict (dict): The model input dict which include

Calculate losses from a batch of inputs and data samples. Args: batch_inputs_dict (dict): The model input dict which include

Loss function. Args: all_layers_cls_scores (Tensor): Classification scores of all decoder layers, has shape (num_

Loss function about feature. Args: center_preds (list[list[Tensor]]): Centerness predictions for all scenes. The

Loss function about feature. Args: center_preds (list[list[Tensor]]): Centerness predictions for all scenes. The

Compared to the ``stack_batch`` in `mmengine.model.utils`, multiview_img_stack_batch further handle the multiview images. See diff of padded_

Non-Maximum Suppression for 3D Euler boxes. Additionally, only the top-k boxes will be kept for each category to avoid redundant boxes in the

Non-Maximum Suppression for 3D Euler boxes. Additionally, only the top-k boxes will be kept for each category to avoid redundant boxes in the

Normalizes a numpy array of points.

Normalizes a numpy array of points.

Obtain multi-level features for each point. Args: img_feats (List[Tensor]): Multi-scale image features produced b

Process the raw data info. The only difference with it in `Det3DDataset` is the specific process for `axis_align_matrix'. Ar

Process the raw data info. The only difference with it in `Det3DDataset` is the specific process for `axis_align_matrix'. Ar

Parse information from a line in a requirements text file.

Predict results from a batch of inputs and data samples with post- processing. Args: batch_inputs_dict (dict): The model

Predict results from a batch of inputs and data samples with post- processing. Args: batch_inputs_dict (dict): The model

Predict results from a batch of inputs and data samples with post- processing. Args: batch_inputs_dict (dict): The model

Predict results from a batch of inputs and data samples with post- processing. Args: batch_inputs_dict (dict): The model

Transform a batch of output features extracted from the head into bbox results. Args: all_layers_cls_scores (Tensor): Cl

Generate boxes for all scenes. Args: center_preds (list[list[Tensor]]): Centerness predictions for all scenes.

Generate boxes for all scenes. Args: center_preds (list[list[Tensor]]): Centerness predictions for all scenes.

Process the files in a directory with parallel support.

This function will be processed after metainfos from ann_file and config are combined.

This function will be processed after metainfos from ann_file and config are combined.

Split projection matrix of KITTI. Note: This function is for KITTI only. P = C @ [R|T] C is upper triangular matrix, so we need

Get voxel coordinates hash for np.unique. Args: x (np.ndarray): The voxel coordinates of points, Nx3. Returns:

Recover input type to original array type. Args: input_array (np.ndarray or torch.Tensor): Input array. Returns:

Reduce loss as specified. Args: loss (Tensor): Elementwise loss tensor. reduction (str): Options are "none", "mean" and "sum".

Render occupancy prediction with continuous ego-centric observations. Args: scene_name (str): Scene name. sta

Render a scene with continuous ego-centric observations. Args: scene_name (str): Scene name. start_cam (str, optional

Calculate the IoU loss (1-IoU) of two sets of rotated bounding boxes. Note that predictions and targets are one-to-one corresponded. Args:

Rotate points by angles according to axis. Args: points (np.ndarray | torch.Tensor | list | tuple ): Points of shape (N, M, 3

Sample features from single level image feature map. Args: img_feats (Tensor): Image feature map in shape (1, C, H, W).

Semantic scene-class affinity loss. Consider probabilities for different categories. Args: pred (Tensor): Prediction maps. s

Show the image on the screen.

Render an ego-centric image view with annotations. Args: scene_name (str): Scene name. camera_name (str): The name of

Shuffle the points. Returns: Tensor: The shuffled index.

Generate grid anchors of a single level feature map. This function is usually called by method ``self.grid_anchors``. Args:

Sparse Quantization for voxel coordinates used in Minkunet. Args: coords (np.ndarray): The voxel coordinates of points, Nx3.

Args: polygon_surfaces (np.ndarray): Polygon surfaces with shape of [num_polygon, max_num_surfaces, max_num_points_of_surfac

Functions to load image. Args: results (dict): Result dict from :class:`mmengine.dataset.BaseDataset`. R

Apply voxelization to point cloud. Args: points (List[Tensor]): Point cloud in one data batch. data_samples: (list[:o

Apply element-wise weight and reduce loss. Args: loss (Tensor): Element-wise loss. weight (Optional[Tensor], optional): Element-w

Calculate nearest 3D IoU. Note: If ``is_aligned`` is ``False``, then it calculates the ious between each bbox of bbox