geowatch.tasks.depth.datasets module

class geowatch.tasks.depth.datasets.WVRgbDataset(*args, **kwargs)

Bases: _CocoTorchDataset

Dataset for depth prediction

Parameters:

dset (kwcoco.CocoDataset) – input dataset to wrap

Example

>>> # xdoctest: +REQUIRES(env:DVC_DPATH)
>>> from geowatch.tasks.depth.datasets import *  # NOQA
>>> import geowatch
>>> import kwcoco
>>> dvc_dpath = geowatch.find_dvc_dpath()
>>> coco_fpath = dvc_dpath / 'Drop2-Aligned-TA1-2022-02-15/data.kwcoco.json'
>>> input_dset = kwcoco.CocoDataset(coco_fpath)
>>> self = WVRgbDataset(input_dset)
>>> # Test that the "include" correctly filter to only WorldView
>>> images = input_dset.images(self.gids)
>>> assert all(s == 'WV' for s in images.lookup('sensor_coarse'))
>>> gid = self.gids[0]
>>> img = self._load(gid)
>>> # xdoctest: +REQUIRES(--show)
>>> import kwplot
>>> kwplot.autompl()
>>> kwplot.imshow(kwarray.normalize(img))
>>> kwplot.show_if_requested()

class geowatch.tasks.depth.datasets.WVSuperRgbDataset(*args, **kwargs)

Bases: _CocoTorchDataset