# Copyright (c) Alibaba, Inc. and its affiliates.
import numpy as np
from torch.utils.data.dataset import ConcatDataset as _ConcatDataset
from easycv.datasets.registry import DATASETS
[docs]@DATASETS.register_module
class ConcatDataset(_ConcatDataset):
"""A wrapper of concatenated dataset.
Same as :obj:`torch.utils.data.dataset.ConcatDataset`, but
concat the group flag for image aspect ratio.
Args:
datasets (list[:obj:`Dataset`]): A list of datasets.
"""
[docs] def __init__(self, datasets):
super(ConcatDataset, self).__init__(datasets)
self.CLASSES = datasets[0].CLASSES
if hasattr(datasets[0], 'flag'):
flags = []
for i in range(0, len(datasets)):
flags.append(datasets[i].flag)
self.flag = np.concatenate(flags)
[docs]@DATASETS.register_module
class RepeatDataset(object):
"""A wrapper of repeated dataset.
The length of repeated dataset will be `times` larger than the original
dataset. This is useful when the data loading time is long but the dataset
is small. Using RepeatDataset can reduce the data loading time between
epochs.
Args:
dataset (:obj:`Dataset`): The dataset to be repeated.
times (int): Repeat times.
"""
[docs] def __init__(self, dataset, times):
self.dataset = dataset
self.times = times
self.CLASSES = dataset.CLASSES
if hasattr(self.dataset, 'flag'):
self.flag = np.tile(self.dataset.flag, times)
self._ori_len = len(self.dataset)
def __getitem__(self, idx):
return self.dataset[idx % self._ori_len]
def __len__(self):
return self.times * self._ori_len