dummy_datasets.py

Create dummy datasets.

The datasets will have the same directory and files structure as the original ones, but only contain a single sample for each directory. The samples are just randomly generated noise.

The main purpose of this script is to be used with tests. But it can also be useful to visualize the structure of a dataset without having to download it.

ptlflow.utils.dummy_datasets.write_flying_chairs(root_dir: str | Path, img_size: Tuple[int, int] = (384, 512)) → None

Generate a dummy version of the Flying Chairs dataset.

The original dataset is available at:

https://lmb.informatik.uni-freiburg.de/resources/datasets/FlyingChairs.en.htm

Parameters:

root_dirUnion[str, Path]

Path to the directory where the dummy dataset will be created.

img_sizeTuple[int, int], default (384, 512)

The size of the images inside of this dataset.

ptlflow.utils.dummy_datasets.write_flying_chairs2(root_dir: str | Path, img_size: Tuple[int, int] = (384, 512)) → None

Generate a dummy version of the Flying Chairs 2 dataset.

The original dataset is available at:

https://lmb.informatik.uni-freiburg.de/resources/datasets/FlyingChairs.en.html

Parameters:

root_dirUnion[str, Path]

Path to the directory where the dummy dataset will be created.

img_sizeTuple[int, int], default (384, 512)

The size of the images inside of this dataset.

ptlflow.utils.dummy_datasets.write_hd1k(root_dir: str | Path, img_size: Tuple[int, int] = (1080, 2560)) → None

Generate a dummy version of the HD1K dataset.

The original dataset is available at:

http://hci-benchmark.org/

Parameters:

root_dirUnion[str, Path]

Path to the directory where the dummy dataset will be created.

img_sizeTuple[int, int], default (1080, 2560)

The size of the images inside of this dataset.

ptlflow.utils.dummy_datasets.write_kitti(root_dir: str | Path, img_size: Tuple[int, int] = (375, 1242)) → None

Generate a dummy version of the KITTI 2012 and 2015 datasets.

The original datasets are available at:

http://www.cvlibs.net/datasets/kitti/eval_stereo_flow.php?benchmark=flow

http://www.cvlibs.net/datasets/kitti/eval_scene_flow.php?benchmark=flow

Parameters:

root_dirUnion[str, Path]

Path to the directory where the dummy dataset will be created.

img_sizeTuple[int, int], default (375, 1242)

The size of the images inside of this dataset.

ptlflow.utils.dummy_datasets.write_sintel(root_dir: str | Path, img_size: Tuple[int, int] = (436, 1024)) → None

Generate a dummy version of the MPI Sintel dataset.

The original dataset is available at:

http://sintel.is.tue.mpg.de/

Parameters:

root_dirUnion[str, Path]

Path to the directory where the dummy dataset will be created.

img_sizeTuple[int, int], default (436, 1024)

The size of the images inside of this dataset.

ptlflow.utils.dummy_datasets.write_things(root_dir: str | Path, img_size: Tuple[int, int] = (540, 960)) → None

Generate a dummy version of the Flying Things 3D dataset.

The original dataset is available at:

https://lmb.informatik.uni-freiburg.de/resources/datasets/SceneFlowDatasets.en.html

Parameters:

root_dirUnion[str, Path]

Path to the directory where the dummy dataset will be created.

img_sizeTuple[int, int], default (540, 960)

The size of the images inside of this dataset.

ptlflow.utils.dummy_datasets.write_things_subset(root_dir: str | Path, img_size: Tuple[int, int] = (540, 960)) → None

Generate a dummy version of the Flying Things 3D subset dataset.

The original dataset is available at:

https://lmb.informatik.uni-freiburg.de/resources/datasets/SceneFlowDatasets.en.html

Parameters:

root_dirUnion[str, Path]

Path to the directory where the dummy dataset will be created.

img_sizeTuple[int, int], default (540, 960)

The size of the images inside of this dataset.