open3d.ml.torch.vis.DatasetModel

class open3d.ml.torch.vis.DatasetModel(dataset, split, indices)

The class used to manage a dataset model.

Parameters

• dataset – The 3D ML dataset to use. You can use the base dataset, sample datasets , or a custom dataset.

• split – A string identifying the dataset split that is usually one of ‘training’, ‘test’, ‘validation’, or ‘all’.

• indices – The indices to be used for the datamodel. This may vary based on the split used.

class BoundingBoxData(name, boxes)

The class to define a bounding box that is used to describe the target location.

Parameters

• name – The name of the pointcloud array.

• boxes – The array of pointcloud that define the bounding box.

__init__(name, boxes)

Initialize self. See help(type(self)) for accurate signature.

__init__(dataset, split, indices)

Initialize self. See help(type(self)) for accurate signature.

calc_bounds_for(name)

Calculate the bounds for a pointcloud.

create_point_cloud(data)

Create a point cloud based on the data provided.

The data should include name and points.

get_attr(name, attr_name)

Get an attribute from data based on the name passed.

get_attr_minmax(attr_name, channel)

Get the minimum and maximum for an attribute.

get_attr_shape(name, attr_name)

Get a shape from data based on the name passed.

get_available_attrs(names)

Get a list of attributes based on the name.

is_loaded(name)

Check if the data is loaded.

load(name, fail_if_no_space=False)

Check if data is not loaded, and then load the data.

unload(name)

Unload the data (if it was loaded earlier).

bounding_box_prefix = 'Bounding Boxes/'