import bisect
import math
from collections import defaultdict
import numpy as np
from torch.utils.data.dataset import ConcatDataset as _ConcatDataset
from .builder import DATASETS
@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.
"""
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)
def get_cat_ids(self, idx):
"""Get category ids of concatenated dataset by index.
Args:
idx (int): Index of data.
Returns:
list[int]: All categories in the image of specified index.
"""
if idx < 0:
if -idx > len(self):
raise ValueError(
'absolute value of index should not exceed dataset length')
idx = len(self) + idx
dataset_idx = bisect.bisect_right(self.cumulative_sizes, idx)
if dataset_idx == 0:
sample_idx = idx
else:
sample_idx = idx - self.cumulative_sizes[dataset_idx - 1]
return self.datasets[dataset_idx].get_cat_ids(sample_idx)
@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.
"""
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 get_cat_ids(self, idx):
"""Get category ids of repeat dataset by index.
Args:
idx (int): Index of data.
Returns:
list[int]: All categories in the image of specified index.
"""
return self.dataset.get_cat_ids(idx % self._ori_len)
def __len__(self):
"""Length after repetition."""
return self.times * self._ori_len
# Modified from https://github.com/facebookresearch/detectron2/blob/41d475b75a230221e21d9cac5d69655e3415e3a4/detectron2/data/samplers/distributed_sampler.py#L57 # noqa
@DATASETS.register_module()
class ClassBalancedDataset(object):
"""A wrapper of repeated dataset with repeat factor.
Suitable for training on class imbalanced datasets like LVIS. Following
the sampling strategy in the `paper <https://arxiv.org/abs/1908.03195>`_,
in each epoch, an image may appear multiple times based on its
"repeat factor".
The repeat factor for an image is a function of the frequency the rarest
category labeled in that image. The "frequency of category c" in [0, 1]
is defined by the fraction of images in the training set (without repeats)
in which category c appears.
The dataset needs to instantiate :func:`self.get_cat_ids` to support
ClassBalancedDataset.
The repeat factor is computed as followed.
1. For each category c, compute the fraction # of images
that contain it: :math:`f(c)`
2. For each category c, compute the category-level repeat factor:
:math:`r(c) = max(1, sqrt(t/f(c)))`
3. For each image I, compute the image-level repeat factor:
:math:`r(I) = max_{c in I} r(c)`
Args:
dataset (:obj:`CustomDataset`): The dataset to be repeated.
oversample_thr (float): frequency threshold below which data is
repeated. For categories with ``f_c >= oversample_thr``, there is
no oversampling. For categories with ``f_c < oversample_thr``, the
degree of oversampling following the square-root inverse frequency
heuristic above.
"""
def __init__(self, dataset, oversample_thr):
self.dataset = dataset
self.oversample_thr = oversample_thr
self.CLASSES = dataset.CLASSES
repeat_factors = self._get_repeat_factors(dataset, oversample_thr)
repeat_indices = []
for dataset_index, repeat_factor in enumerate(repeat_factors):
repeat_indices.extend([dataset_index] * math.ceil(repeat_factor))
self.repeat_indices = repeat_indices
flags = []
if hasattr(self.dataset, 'flag'):
for flag, repeat_factor in zip(self.dataset.flag, repeat_factors):
flags.extend([flag] * int(math.ceil(repeat_factor)))
assert len(flags) == len(repeat_indices)
self.flag = np.asarray(flags, dtype=np.uint8)
def _get_repeat_factors(self, dataset, repeat_thr):
"""Get repeat factor for each images in the dataset.
Args:
dataset (:obj:`CustomDataset`): The dataset
repeat_thr (float): The threshold of frequency. If an image
contains the categories whose frequency below the threshold,
it would be repeated.
Returns:
list[float]: The repeat factors for each images in the dataset.
"""
# 1. For each category c, compute the fraction # of images
# that contain it: f(c)
category_freq = defaultdict(int)
num_images = len(dataset)
for idx in range(num_images):
cat_ids = set(self.dataset.get_cat_ids(idx))
for cat_id in cat_ids:
category_freq[cat_id] += 1
for k, v in category_freq.items():
category_freq[k] = v / num_images
# 2. For each category c, compute the category-level repeat factor:
# r(c) = max(1, sqrt(t/f(c)))
category_repeat = {
cat_id: max(1.0, math.sqrt(repeat_thr / cat_freq))
for cat_id, cat_freq in category_freq.items()
}
# 3. For each image I, compute the image-level repeat factor:
# r(I) = max_{c in I} r(c)
repeat_factors = []
for idx in range(num_images):
cat_ids = set(self.dataset.get_cat_ids(idx))
repeat_factor = max(
{category_repeat[cat_id]
for cat_id in cat_ids})
repeat_factors.append(repeat_factor)
return repeat_factors
def __getitem__(self, idx):
ori_index = self.repeat_indices[idx]
return self.dataset[ori_index]
def __len__(self):
"""Length after repetition."""
return len(self.repeat_indices)
