vp_suite.datasets.physics101

class Physics101Dataset(split, **dataset_kwargs)

Bases: vp_suite.base.base_dataset.VPDataset

Dataset class for the Videos of the dataset “Physics 101”, as encountered in “Physics 101: Learning Physical Object Properties from Unlabeled Videos” by Wu et al. (http://phys101.csail.mit.edu/papers/phys101_bmvc.pdf).

Each sequence depicts object-centered physical properties by showing objects of various materials and apperances in different physical scenarios such as sliding down a ramp or bouncing off a flat surface.

ACTION_SIZE: int = 0

The size of the action vector per frame (If the dataset provides no actions, this value is 0).

AVAILABLE_CAMERAS = ['Camera_1', 'Camera_2', 'Kinect_RGB_1']

Available cameras/image sources.

AVAILABLE_SUBSEQ = ['start', 'middle', 'end']

Available (sub-)sequence extraction position identifiers.

DATASET_FRAME_SHAPE: (<class 'int'>, <class 'int'>, <class 'int'>) = (1080, 1920, 3)

Shape of a single frame in the dataset (height, width, channels).

DEFAULT_DATA_DIR: pathlib.Path = PosixPath('/home/runner/work/vp-suite/vp-suite/vp-suite-data/data/phys101')

The default save location of the dataset files.

IS_DOWNLOADABLE: str = 'Yes'

A string identifying whether the dataset can be (freely) downloaded.

MIN_SEQ_LEN: int = 16

The minimum sequence length provided by the dataset.

NAME: str = 'Physics 101'

The dataset’s name.

REFERENCE: str = 'http://phys101.csail.mit.edu/'

The reference (publication) where the original dataset is introduced.

__init__(split, **dataset_kwargs)

Initializes the dataset loader by determining its split and extracting and processing all dataset attributes from the parameters given in dataset_kwargs.

Parameters

• split (str) – The dataset’s split identifier (i.e. whether it’s a training/validation/test dataset)

• **dataset_kwargs (Any) – Optional dataset arguments for image transformation, value_range, splitting etc.

camera = 'Kinect_RGB_1'

Which camera to use from the dataset.

download_and_prepare_dataset()

Downloads the specific dataset, prepares it for the video prediction task (if needed) and stores it in a default location in the ‘data/’ folder. Implemented by the derived dataset classes.

subseq = 'middle'

“start” starts from the first frame, “end” ends at the last frame and “middle” lies exactly in between.

Type

Where to extract the sequence from

trainval_test_seed = 1612

The random seed used to separate training/validation and testing data. Value from the ‘Noether Networks’ code

trainval_to_test_ratio = 0.8

The ratio of files that will be training/validation data (rest will be test data).