142 lines
5.0 KiB
Python

# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
from __future__ import unicode_literals
import numpy as np
from ppcls.data.preprocess.ops.fmix import sample_mask
class BatchOperator(object):
""" BatchOperator """
def __init__(self, *args, **kwargs):
pass
def _unpack(self, batch):
""" _unpack """
assert isinstance(batch, list), \
'batch should be a list filled with tuples (img, label)'
bs = len(batch)
assert bs > 0, 'size of the batch data should > 0'
#imgs, labels = list(zip(*batch))
imgs = []
labels = []
for item in batch:
imgs.append(item[0])
labels.append(item[1])
return np.array(imgs), np.array(labels), bs
def __call__(self, batch):
return batch
class MixupOperator(BatchOperator):
"""Mixup and Cutmix operator"""
def __init__(self,
mixup_alpha: float=1.,
cutmix_alpha: float=0.,
switch_prob: float=0.5):
"""Build Mixup operator
Args:
mixup_alpha (float, optional): The parameter alpha of mixup, mixup is active if > 0. Defaults to 1..
cutmix_alpha (float, optional): The parameter alpha of cutmix, cutmix is active if > 0. Defaults to 0..
switch_prob (float, optional): The probability of switching to cutmix instead of mixup when both are active. Defaults to 0.5.
Raises:
Exception: The value of parameters are illegal.
"""
if mixup_alpha <= 0 and cutmix_alpha <= 0:
raise Exception(
f"At least one of parameter alpha of Mixup and Cutmix is greater than 0. mixup_alpha: {mixup_alpha}, cutmix_alpha: {cutmix_alpha}"
)
self._mixup_alpha = mixup_alpha
self._cutmix_alpha = cutmix_alpha
self._switch_prob = switch_prob
def _mixup(self, imgs, labels, bs):
idx = np.random.permutation(bs)
lam = np.random.beta(self._mixup_alpha, self._mixup_alpha)
lams = np.array([lam] * bs, dtype=np.float32)
imgs = lam * imgs + (1 - lam) * imgs[idx]
return list(zip(imgs, labels, labels[idx], lams))
def _rand_bbox(self, size, lam):
""" _rand_bbox """
w = size[2]
h = size[3]
cut_rat = np.sqrt(1. - lam)
cut_w = int(w * cut_rat)
cut_h = int(h * cut_rat)
# uniform
cx = np.random.randint(w)
cy = np.random.randint(h)
bbx1 = np.clip(cx - cut_w // 2, 0, w)
bby1 = np.clip(cy - cut_h // 2, 0, h)
bbx2 = np.clip(cx + cut_w // 2, 0, w)
bby2 = np.clip(cy + cut_h // 2, 0, h)
return bbx1, bby1, bbx2, bby2
def _cutmix(self, imgs, labels, bs):
idx = np.random.permutation(bs)
lam = np.random.beta(self._cutmix_alpha, self._cutmix_alpha)
bbx1, bby1, bbx2, bby2 = self._rand_bbox(imgs.shape, lam)
imgs[:, :, bbx1:bbx2, bby1:bby2] = imgs[idx, :, bbx1:bbx2, bby1:bby2]
lam = 1 - (float(bbx2 - bbx1) * (bby2 - bby1) /
(imgs.shape[-2] * imgs.shape[-1]))
lams = np.array([lam] * bs, dtype=np.float32)
return list(zip(imgs, labels, labels[idx], lams))
def __call__(self, batch):
imgs, labels, bs = self._unpack(batch)
if np.random.rand() < self._switch_prob:
return self._cutmix(imgs, labels, bs)
else:
return self._mixup(imgs, labels, bs)
class CutmixOperator(BatchOperator):
def __init__(self, **kwargs):
raise Exception(
f"\"CutmixOperator\" has been deprecated. Please use MixupOperator with \"cutmix_alpha\" and \"switch_prob\" to enable Cutmix. Refor to doc for details."
)
class FmixOperator(BatchOperator):
""" Fmix operator """
def __init__(self, alpha=1, decay_power=3, max_soft=0., reformulate=False):
self._alpha = alpha
self._decay_power = decay_power
self._max_soft = max_soft
self._reformulate = reformulate
def __call__(self, batch):
imgs, labels, bs = self._unpack(batch)
idx = np.random.permutation(bs)
size = (imgs.shape[2], imgs.shape[3])
lam, mask = sample_mask(self._alpha, self._decay_power, \
size, self._max_soft, self._reformulate)
imgs = mask * imgs + (1 - mask) * imgs[idx]
return list(zip(imgs, labels, labels[idx], [lam] * bs))