Final categories used in the generated dataset Minor tweaks from fine categories

Return pairwise prior for given category_key

Create the actual multi-channel representation. Which is a N x img_size x img_size tensor. See the paper for more information

Parameters ---------- source, dest (String): location of source/dest relative (should be) to utils.get_data_root_dir() stats (bool):

Save the json file, see save_top_down_view

Save the top down view of the current room Parameters ---------- save_dir (String): location to be saved fin

Add a new object to the composite. Height map, category, and angle of rotation are all the information required.

Create a initial composite that only contains the floor, wall, doors and windows. See RenderedComposite for how to add more o

Since some models in the dataset are not aligned in the way we want Generate matrix that realign them

Get the top-down render of the object

Parameters ---------- t(Matrix or None): transformation matrix of the object if None, then returns room projectio

Add a new object into the current room

Construct Transform from matrix m :param m: 4x4 affine transformation matrix :return: rotation (quaternion), translation, sca

Return all current contacts. When include_collision_with_statics is true, include contacts with static bodies

Generates projection matrices specific to a room, need to be room-specific since every room is located in a different position,

Get the transformation matrix Used to render the object and to save in json files

Render a cropped top-down view of object Parameters ---------- o (list[triple]): object to be rendered, repr

Sometimes we need to create a composite to test if some objects should be added, without actually fixing the object to the sc

Create a pickled version of the dataset from the json files Parameters ---------- dest (string, optional): directory to save to

Return canonical alignment dimensions of model *in meters*

Return record with distance between closest points between pair of nodes if within max_distance or None.

Get the json of the current room, plus additional_nodes Parameters ---------- additional_nodes (list[SynthNode]): ob

Get a set of nodes from the room which satisfies a certain criteria Parameters ---------- filters (list[node_filter]

parse .obj objects and save them to pickle files

Render a full-sized top-down view of the object, see render_object

From https://stackoverflow.com/questions/5081657/how-do-i-prevent-a-c-shared-library-to-print-on-stdout-in-python Suppress C warnings