Return the size of the point cloud.

Make `p` listy and the same length as `q`.

Set `val` inside the optimizer dictionary at `key`.

Set weight decay and step optimizer.

Args: code_weights: (#codes) float list if not None. Code-wise weights.

nanmean compatible with generators.

Open raw scan and fill in attributes.

Read a hyperparameter `key` in the optimizer dictionary.

1x1 convolution

Create an `optim.Optimizer` from `opt_func` with `lr`. Set lr on `layer_groups`.

Flattens predictions in the batch

Arguments: - strategy: MixTeacher strategies. - image: original image, size: [6, H, W]. - label: original

convert 2d range feature map to points feature

Reset scan members.

Clear optimizer gradients.

3x3 convolution with padding

Flattens predictions in the batch (binary case) Remove labels equal to 'ignore'

Open raw scan and fill in attributes.

args: pf: N x C point features pxpy: N x 3 batch_id,px,py px range: -1~1 py range: -1~1 b: batch_

Visualize the points cloud. Args: save_path (str, optional): path to save image. Default: None.

Split the layers in `layer_groups` into batchnorm (`bn_types`) and non-batchnorm groups.

Draw points on visualizer. Args: points (numpy.array | torch.tensor, shape=[N, 3+C]): points to visualize. vis (:obj:`

Draw points on visualizer. Args: points (numpy.array | torch.tensor, shape=[N, 3+C]): points to visualize. vis (:obj:`

Set config keys via list (e.g., from command line).

Modified from the codes of Waymo Open Dataset. Convert range images to point cloud. Args: frame: open dataset frame range

Convert segmentation labels from range images to point clouds. Args: frame: open dataset frame range_images: A dict of {laser_name, [

Flattens predictions in the batch

Flattens predictions in the batch

Generates one sample of data

Generates one sample of data

Generates one sample of data

Generates one sample of data

Generates one sample of data

Generates one sample of data