Print a log message. Args: msg (str): The message to be logged. logger (logging.Logger | str | None): The logger to be used.

Initialize the detectron2 logger and set its verbosity level to "INFO". Args: output (str): a file name or a directory to save log. I

:param mlp: list of channel sizes

Create an optimizer. Args: model (nn.Module): model containing parameters to optimize NAME: name of optimizer to create l

Build a dataset transform for a certrain split, defined by `datatransforms_cfg`.

Input: xyz: pointcloud data, [N, D] npoint: number of samples Return: centroids: sampled pointcloud index, [npoint, D

Function to create checkpoint folder. Args: cfg: configuration dict cfg.root_dir: the root dir for saving log files. exp_n

Args: points [point]: (B, N, C) idx [sample index data]: (B, S) Return: new_points [indexed points data]: (B, S,

Input: points: input points data, [B, N, C] idx: sample index data, [B, S] Return: new_points:, indexed points data,

Input: points: input points data, [B, N, C] idx: sample index data, [B, S] Return: new_points:, indexed points data,

GetItemInt */

ParseKeywords */

1x1 convolution

3x3 convolution with padding

Input: xyz: pointcloud data, [B, N, 3] npoint: number of samples Return: centroids: sampled pointcloud index, [B, npo

Frees the previously-built index. Automatically called within buildIndex(). */

Frees the previously-built index. Automatically called within * buildIndex(). */

Frees the previously-built index. Automatically called within * buildIndex(). */

Get the registry record. Args: key (str): The class name in string format. Returns: class: The corresponding c

Gets all parameters in all param_groups with gradients

Input: points: input points data, [B, N, C] idx: sample index data, [B, S] Return: new_points:, indexed points data,

Register a module. A record will be added to `self._module_dict`, whose key is the class name or the specified name, and value is the

Function to resume the exp folder from the checkpoint folder. Args: cfg pretrained_path: the path to the pretrained model Retu

using a part of original point cloud as input to save memory. Args: model (_type_): _description_ test_loader (_type_): _descripti

input: known_xyz: (m, 3), unknown_xyz: (n, 3), feat: (m, c), offset: (b), new_offset: (b) output: (n, c)