List all scenes in the dataset.

Load annotations from an annotation file named as ``self.ann_file`` If the annotation file does not follow `OpenMMLab 2.0 format dataset

Load annotations from an annotation file named as ``self.ann_file`` If the annotation file does not follow `OpenMMLab 2.0 format dataset

Perform forward propagation and loss calculation of the detection head on the queries of the upstream network. Args: hidd

Loss function.. Args: x (Tuple[Tensor]): Multi-level features. batch_data_samples (`SampleList`): Batch of data sampl

Perform forward propagation and loss calculation of the detection head on the features of the upstream network. Args: x (

Perform forward propagation and loss calculation of the detection head on the features of the upstream network. Args: x (

Loss function for outputs from a single decoder layer of a single feature level. Args: cls_scores (Tensor): Box score log

Convert minmax box to corners2d. Args: minmax_box (np.ndarray, shape=[N, dims]): minmax boxes. Returns: np.ndarray: 2d corne

This is a post-processing function on the bboxes from Mono-3D task. If we want to perform projection visualization, we need to: 1. rotate

Tensor: A tensor of 2D BEV box of each box without rotation.

Inner wrapper for the arguments.

Create a new point object with data. The new point and its tensor has the similar properties as self and self.tensor, respectively.

Find boxes that are non-empty. A box is considered empty if either of its side is no larger than threshold. Args:

int: Total number of base anchors in a feature grid.

int: Number of feature levels that the generator is applied to.

Find all boxes in which each point is. Args: points (torch.Tensor): Points in shape (1, M, 3) or (M, 3), 3 dimens

Find the box in which each point is. Args: points (torch.Tensor): Points in shape (1, M, 3) or (M, 3), 3 dimensio

Check points is in 2d convex polygons. True when point in polygon. Args: points (np.ndarray): Input points with the shape of [num_points,

Check points in rotated bbox and return indices. Note: This function is for counterclockwise boxes. Args: points (np.ndarray

Perform forward propagation and loss calculation of the detection head on the queries of the upstream network. Args: hidd

Predict/Inference function. Args: x (Tuple[Tensor]): Multi-level features. batch_data_samples (`SampleList`): Batch o

Perform forward propagation of the 3D detection head and predict detection results on the features of the upstream network. Args:

Perform forward propagation of the 3D detection head and predict detection results on the features of the upstream network. Args:

Process one batch of data samples and predictions. The processed results should be stored in ``self.results``, which will be used to

Process one batch of data samples and predictions. The processed results should be stored in ``self.results``, which will be used to

Process one batch of data samples and predictions. The processed results should be stored in ``self.results``, which will be used to

Process one batch of data samples and predictions. The processed results should be stored in ``self.results``, which will be used to

Process single ScanNet scene. Extract RGB images, poses and camera intrinsics.

Process single 3Rscan scene.

convert rotated bbox to nearest 'standing' or 'lying' bbox. Args: rbboxes (np.ndarray): Rotated bboxes with shape of (N, 5(x,

Removes this line from the visualizer.

Removes this line from the visualizer.

Remove points which are outside of image. Note: This function is for KITTI only. Args: points (np.ndarray, shape=[N, 3+dims]

Render occupancy with continuous ego-centric observations. Args: scene_name (str): Scene name. start_cam (str, option

Render the occupancy annotation of a given scene. Args: scene_name (str): Scene name.

Render a given scene with open3d. Args: scene_name (str): Scene name. render_box (bool): Whether to render the box in

Rotate boxes with points (optional) with the given angle or rotation matrix. Args: angle (float | torch.Tensor | np.ndarr

Rotate boxes with points (optional) with the given angle or rotation matrix. Args: angle (Tensor or np.ndarray or float):

Rotate points with a single angle. Args: points (np.ndarray, shape=[N, 3]]): angle (np.ndarray, shape=[1]]): axis (int, o

Save the colormap to the output path. Args: out_file (str): Output path for saving the colormap.

Scale the box with horizontal and vertical scaling factors. Args: scale_factors (float): Scale factors to scale the boxes.

Scale the box with horizontal and vertical scaling factors. Args: scale_factors (float): Scale factors to scale the boxes.

Show the info of the given scene. Args: scene_name (str): Scene name. Returns: dict or None: Dict of scene i

torch.Size: Shape of points.

torch.Size: Shape of boxes.

Tensor: A vector with top height of each box in shape (N, ).

Call function to flip points, values in the ``bbox3d_fields`` and also flip 2D image and its annotations. Args: input_dic

Private function to rotate, scale and translate bounding boxes and points. Args: input_dict (dict): Result dict from load

Function to load multiple types annotations. Args: results (dict): Result dict from :obj:`mmdet3d.CustomDataset`. Return

Call function to normalize color of points. Args: input_dict (dict): Result dict containing point clouds data. Returns:

Transform function to filter points by the range. Args: input_dict (dict): Result dict from loading pipeline. Returns:

Transform function. Args: results (dict): Result dict from loading pipeline. Returns: dict: output dict afte

Method to pack the input data. when the value in this dict is a list, it usually is in Augmentations Testing. Args: resul

Translate boxes with the given translation vector. Args: trans_vector (Tensor or np.ndarray): Translation vector of size

Visualize the 3D scene with 3D boxes. Args: data_samples (list[:obj:`Det3DDataSample`]): The output of the model.

Tensor: A vector with volume of each box in shape (N, ).

Obtain image features using points. Args: voxel_features (Tensor): 1 x C x Nx x Ny x Nz voxel features. voxel_range (List[float])

Create a weighted version of a given loss function. To use this decorator, the loss function must have the signature like `loss_func(pred, ta

Whether the detector has a 2D backbone.

Whether the detector has a 2D backbone.

Whether the detector has a 2D backbone.

Whether the detector has a 2D backbone.

Whether the detector has a 2D backbone.

Whether the detector has a 3D neck.

Whether the detector has a 3D neck.

Whether the detector has a 3D neck.

Whether the detector has a 3D neck.

Whether the detector has a 3D neck.

Whether the detector has a 2D backbone.

Args: pred (Tensor): The prediction. target (Tensor): Target bboxes. weight (Optional[Tensor], optional):

Convert a rotated boxes in XYWHR format to XYXYR format. Args: boxes_xywhr (Tensor or np.ndarray): Rotated boxes in XYWHR format. Re

Tensor: A vector with yaw of each box in shape (N, ).

Transform global yaw to local yaw (alpha in kitti) in camera coordinates, ranges from -pi to pi. Args: yaw (Tensor): A vector with lo