Shortcuts

Source code for mmaction.models.utils.blending_utils

# Copyright (c) OpenMMLab. All rights reserved.
from abc import ABCMeta, abstractmethod
from functools import partial
from typing import List, Optional, Tuple, Union

import numpy as np
import torch
import torch.nn.functional as F
from mmengine.utils import digit_version
from torch.distributions.beta import Beta

from mmaction.registry import MODELS
from mmaction.utils import SampleList

if digit_version(torch.__version__) < digit_version('1.8.0'):
    floor_div = torch.floor_divide
else:
    floor_div = partial(torch.div, rounding_mode='floor')

__all__ = ['BaseMiniBatchBlending', 'MixupBlending', 'CutmixBlending']


[docs]class BaseMiniBatchBlending(metaclass=ABCMeta): """Base class for Image Aliasing. Args: num_classes (int): Number of classes. """ def __init__(self, num_classes: int) -> None: self.num_classes = num_classes
[docs] @abstractmethod def do_blending(self, imgs: torch.Tensor, label: torch.Tensor, **kwargs) -> Tuple: """Blending images process.""" raise NotImplementedError
def __call__(self, imgs: torch.Tensor, batch_data_samples: SampleList, **kwargs) -> Tuple: """Blending data in a mini-batch. Images are float tensors with the shape of (B, N, C, H, W) for 2D recognizers or (B, N, C, T, H, W) for 3D recognizers. Besides, labels are converted from hard labels to soft labels. Hard labels are integer tensors with the shape of (B, ) and all of the elements are in the range [0, num_classes - 1]. Soft labels (probability distribution over classes) are float tensors with the shape of (B, num_classes) and all of the elements are in the range [0, 1]. Args: imgs (torch.Tensor): Model input images, float tensor with the shape of (B, N, C, H, W) or (B, N, C, T, H, W). batch_data_samples (List[:obj:`ActionDataSample`]): The batch data samples. It usually includes information such as `gt_label`. Returns: mixed_imgs (torch.Tensor): Blending images, float tensor with the same shape of the input imgs. batch_data_samples (List[:obj:`ActionDataSample`]): The modified batch data samples. ``gt_label`` in each data sample are converted from a hard label to a blended soft label, float tensor with the shape of (num_classes, ) and all elements are in range [0, 1]. """ label = [x.gt_label for x in batch_data_samples] # single-label classification if label[0].size(0) == 1: label = torch.tensor(label, dtype=torch.long).to(imgs.device) one_hot_label = F.one_hot(label, num_classes=self.num_classes) # multi-label classification else: one_hot_label = torch.stack(label) mixed_imgs, mixed_label = self.do_blending(imgs, one_hot_label, **kwargs) for label_item, sample in zip(mixed_label, batch_data_samples): sample.set_gt_label(label_item) return mixed_imgs, batch_data_samples
[docs]@MODELS.register_module() class MixupBlending(BaseMiniBatchBlending): """Implementing Mixup in a mini-batch. This module is proposed in `mixup: Beyond Empirical Risk Minimization <https://arxiv.org/abs/1710.09412>`_. Code Reference https://github.com/open-mmlab/mmclassification/blob/master/mmcls/models/utils/mixup.py # noqa Args: num_classes (int): The number of classes. alpha (float): Parameters for Beta distribution. """ def __init__(self, num_classes: int, alpha: float = .2) -> None: super().__init__(num_classes=num_classes) self.beta = Beta(alpha, alpha)
[docs] def do_blending(self, imgs: torch.Tensor, label: torch.Tensor, **kwargs) -> Tuple: """Blending images with mixup. Args: imgs (torch.Tensor): Model input images, float tensor with the shape of (B, N, C, H, W) or (B, N, C, T, H, W). label (torch.Tensor): One hot labels, integer tensor with the shape of (B, num_classes). Returns: tuple: A tuple of blended images and labels. """ assert len(kwargs) == 0, f'unexpected kwargs for mixup {kwargs}' lam = self.beta.sample() batch_size = imgs.size(0) rand_index = torch.randperm(batch_size) mixed_imgs = lam * imgs + (1 - lam) * imgs[rand_index, :] mixed_label = lam * label + (1 - lam) * label[rand_index, :] return mixed_imgs, mixed_label
[docs]@MODELS.register_module() class CutmixBlending(BaseMiniBatchBlending): """Implementing Cutmix in a mini-batch. This module is proposed in `CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features <https://arxiv.org/abs/1905.04899>`_. Code Reference https://github.com/clovaai/CutMix-PyTorch Args: num_classes (int): The number of classes. alpha (float): Parameters for Beta distribution. """ def __init__(self, num_classes: int, alpha: float = .2) -> None: super().__init__(num_classes=num_classes) self.beta = Beta(alpha, alpha)
[docs] @staticmethod def rand_bbox(img_size: torch.Size, lam: torch.Tensor) -> Tuple: """Generate a random boudning box.""" w = img_size[-1] h = img_size[-2] cut_rat = torch.sqrt(1. - lam) cut_w = torch.tensor(int(w * cut_rat)) cut_h = torch.tensor(int(h * cut_rat)) # uniform cx = torch.randint(w, (1, ))[0] cy = torch.randint(h, (1, ))[0] bbx1 = torch.clamp(cx - floor_div(cut_w, 2), 0, w) bby1 = torch.clamp(cy - floor_div(cut_h, 2), 0, h) bbx2 = torch.clamp(cx + floor_div(cut_w, 2), 0, w) bby2 = torch.clamp(cy + floor_div(cut_h, 2), 0, h) return bbx1, bby1, bbx2, bby2
[docs] def do_blending(self, imgs: torch.Tensor, label: torch.Tensor, **kwargs) -> Tuple: """Blending images with cutmix. Args: imgs (torch.Tensor): Model input images, float tensor with the shape of (B, N, C, H, W) or (B, N, C, T, H, W). label (torch.Tensor): One hot labels, integer tensor with the shape of (B, num_classes). Returns: tuple: A tuple of blended images and labels. """ assert len(kwargs) == 0, f'unexpected kwargs for cutmix {kwargs}' batch_size = imgs.size(0) rand_index = torch.randperm(batch_size) lam = self.beta.sample() bbx1, bby1, bbx2, bby2 = self.rand_bbox(imgs.size(), lam) imgs[:, ..., bby1:bby2, bbx1:bbx2] = imgs[rand_index, ..., bby1:bby2, bbx1:bbx2] lam = 1 - (1.0 * (bbx2 - bbx1) * (bby2 - bby1) / (imgs.size()[-1] * imgs.size()[-2])) label = lam * label + (1 - lam) * label[rand_index, :] return imgs, label
[docs]@MODELS.register_module() class RandomBatchAugment(BaseMiniBatchBlending): """Randomly choose one batch augmentation to apply. Args: augments (dict | list): configs of batch augmentations. probs (float | List[float] | None): The probabilities of each batch augmentations. If None, choose evenly. Defaults to None. Example: >>> augments_cfg = [ ... dict(type='CutmixBlending', alpha=1., num_classes=10), ... dict(type='MixupBlending', alpha=1., num_classes=10) ... ] >>> batch_augment = RandomBatchAugment(augments_cfg, probs=[0.5, 0.3]) >>> imgs = torch.randn(16, 3, 8, 32, 32) >>> label = torch.randint(0, 10, (16, )) >>> imgs, label = batch_augment(imgs, label) .. note :: To decide which batch augmentation will be used, it picks one of ``augments`` based on the probabilities. In the example above, the probability to use CutmixBlending is 0.5, to use MixupBlending is 0.3, and to do nothing is 0.2. """ def __init__(self, augments: Union[dict, list], probs: Optional[Union[float, List[float]]] = None) -> None: if not isinstance(augments, (tuple, list)): augments = [augments] self.augments = [] for aug in augments: assert isinstance(aug, dict), \ f'blending augment config must be a dict. Got {type(aug)}' self.augments.append(MODELS.build(aug)) self.num_classes = augments[0].get('num_classes') if isinstance(probs, float): probs = [probs] if probs is not None: assert len(augments) == len(probs), \ '``augments`` and ``probs`` must have same lengths. ' \ f'Got {len(augments)} vs {len(probs)}.' assert sum(probs) <= 1, \ 'The total probability of batch augments exceeds 1.' self.augments.append(None) probs.append(1 - sum(probs)) self.probs = probs
[docs] def do_blending(self, imgs: torch.Tensor, label: torch.Tensor, **kwargs) -> Tuple: """Randomly apply batch augmentations to the batch inputs and batch data samples.""" aug_index = np.random.choice(len(self.augments), p=self.probs) aug = self.augments[aug_index] if aug is not None: return aug.do_blending(imgs, label, **kwargs) else: return imgs, label
Read the Docs v: latest
Versions
latest
stable
1.x
0.x
dev-1.x
Downloads
epub
On Read the Docs
Project Home
Builds

Free document hosting provided by Read the Docs.