7e470179b0d874c43ec6d5dba11d7981ad83dff1,chainercv/datasets/sintel/sintel_dataset.py,SintelDataset,__init__,#SintelDataset#Any#Any#,58

Before Change

            data_dir = os.path.join(data_root, "training")
        self.data_dir = data_dir
        self.paths = self._collect_data(data_dir)
        self.keys = self.paths.keys()

        self.mode = mode

    def __len__(self):

After Change

    This class returns tuple of (source, target, flow) as :meth:`get_example`
    is called.
    `source` is the image of the source image and `target` is the image of
    the target image which are both in CHW format. `flow` represents optical
    flow from the source to the target whose shape is :math:`(3, H, W)`.
    :math:`H` and :math:`W` are the height and the width of images.

    Args:
        data_dir (string): Path to the root of the training data. If this is
            :obj:`auto`, this class will automatically download data for you
            under :obj:`$CHAINER_DATASET_ROOT/pfnet/chainercv/sintel`.
    

Instances