CutMix: Regularization Strategy to Train Strong Classifiers With Localizable Features

@article{Yun2019CutMixRS,
  title={CutMix: Regularization Strategy to Train Strong Classifiers With Localizable Features},
  author={Sangdoo Yun and Dongyoon Han and Seong Joon Oh and Sanghyuk Chun and Junsuk Choe and Young Joon Yoo},
  journal={2019 IEEE/CVF International Conference on Computer Vision (ICCV)},
  year={2019},
  pages={6022-6031}
}
Regional dropout strategies have been proposed to enhance the performance of convolutional neural network classifiers. [...] Key Method We therefore propose the CutMix augmentation strategy: patches are cut and pasted among training images where the ground truth labels are also mixed proportionally to the area of the patches. By making efficient use of training pixels and retaining the regularization effect of regional dropout, CutMix consistently outperforms the state-of-the-art augmentation strategies on…Expand
SmoothMix: a Simple Yet Effective Data Augmentation to Train Robust Classifiers
TLDR
Smoothmix is introduced in which blending of images is done based on soft edges and the training labels are computed accordingly, which significantly increases the robustness of a network against image corruption which is validated by the experiments carried out on CIFAR-100-C & ImageNet-C corruption datasets. Expand
Attentive Cutmix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification
TLDR
Attentive CutMix is proposed, a naturally enhanced augmentation strategy based on CutMix that consistently outperforms the baseline CutMix and other methods by a significant margin, and can boost the baseline significantly. Expand
SaliencyMix: A Saliency Guided Data Augmentation Strategy for Better Regularization
TLDR
This work proposes SaliencyMix, a new state-of-the-art top-1 error-reducing model that carefully selects a representative image patch with the help of a saliency map and mixes this indicative patch with a target image that leads the model to learn more appropriate feature representation. Expand
SALIENCYMIX: A SALIENCY GUIDED DATA AUG-
Advanced data augmentation strategies have widely been studied to improve the generalization ability of deep learning models. Regional dropout is one of the popular solutions that guides the model toExpand
Region-based dropout with attention prior for weakly supervised object localization
TLDR
The region-based dropout with attention prior (RDAP) algorithm, which features hyperparameter transferability is proposed, which achieved state-of-the-art localization accuracy on four architectures, namely VGG-GAP, InceptionV3, ResNet-50 SE, and PreResNet-18, and two datasets, namely CUB-200-2011 and ImageNet-1k, with a single set of hyperparameters. Expand
Where to Cut and Paste: Data Regularization with Selective Features
TLDR
A new data augmentation method strategy, called FocusMix, which exploits informative pixels based on proper sampling techniques is proposed, and it is shown that FocusMix results in improvements in performance compared to otherData augmentation methods. Expand
Towards Learning Spatially Discriminative Feature Representations
  • Chaofei Wang, Jiayu Xiao, Yizeng Han, Qisen Yang, Shiji Song, Gao Huang
  • Computer Science
  • ArXiv
  • 2021
TLDR
A novel loss function, termed as CAM-loss, is proposed to constrain the embedded feature maps with the class activation maps (CAMs) which indicate the spatially discriminative regions of an image for particular categories. Expand
ResizeMix: Mixing Data with Preserved Object Information and True Labels
TLDR
ResizeMix shows evident advantages over CutMix and the saliency-guided methods on both image classification and object detection tasks without additional computation cost, which even outperforms most costly search-based automatic augmentation methods. Expand
Regional Patch-Based Feature Interpolation Method for Effective Regularization
TLDR
This paper proposes a regularization method that applies both image manipulation and feature map regularization based on patches that improves the performance of the model and makes it possible to better generalize the model. Expand
E2Net: Excitative-Expansile Learning for Weakly Supervised Object Localization
  • Zhiwei Chen, Liujuan Cao, Yunhang Shen, Feihong Lian, Yongjian Wu, Rongrong Ji
  • Computer Science
  • ACM Multimedia
  • 2021
TLDR
A novel end-to-end Excitation-Expansion network to localize entire objects with only image-level labels, coined as E$^2$Net, which outperforms the previous state-of-the-art WSOL methods and also significantly improves classification performance. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 56 REFERENCES
Improved Regularization of Convolutional Neural Networks with Cutout
TLDR
This paper shows that the simple regularization technique of randomly masking out square regions of input during training, which is called cutout, can be used to improve the robustness and overall performance of convolutional neural networks. Expand
LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop
TLDR
This work proposes to amplify human effort through a partially automated labeling scheme, leveraging deep learning with humans in the loop, and constructs a new image dataset, LSUN, which contains around one million labeled images for each of 10 scene categories and 20 object categories. Expand
Attention-Based Dropout Layer for Weakly Supervised Object Localization
  • Junsuk Choe, Hyunjung Shim
  • Computer Science
  • 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2019
TLDR
An Attention-based Dropout Layer (ADL), which utilizes the self-attention mechanism to process the feature maps of the model to improve the accuracy of WSOL, achieving a new state-of-the-art localization accuracy in CUB-200-2011 dataset. Expand
DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
TLDR
This work addresses the task of semantic image segmentation with Deep Learning and proposes atrous spatial pyramid pooling (ASPP), which is proposed to robustly segment objects at multiple scales, and improves the localization of object boundaries by combining methods from DCNNs and probabilistic graphical models. Expand
Learning Deep Features for Discriminative Localization
In this work, we revisit the global average pooling layer proposed in [13], and shed light on how it explicitly enables the convolutional neural network (CNN) to have remarkable localization abilityExpand
RICAP: Random Image Cropping and Patching Data Augmentation for Deep CNNs
TLDR
A new data augmentation technique called random image cropping and patching (RICAP), which randomly crops four images and patches them to construct a new training image, enriching the variety of training images. Expand
Self-produced Guidance for Weakly-supervised Object Localization
TLDR
Self-produced Guidance (SPG) masks which separate the foreground i.e., the object of interest, from the background to provide the classification networks with spatial correlation information of pixels are proposed. Expand
Adversarial Complementary Learning for Weakly Supervised Object Localization
TLDR
This work mathematically proves that class localization maps can be obtained by directly selecting the class-specific feature maps of the last convolutional layer, which paves a simple way to identify object regions and presents a simple network architecture including two parallel-classifiers for object localization. Expand
Modeling Visual Context is Key to Augmenting Object Detection Datasets
TLDR
This work leverages segmentation annotations to increase the number of object instances present on training data and shows that modeling appropriately the visual context surrounding objects is crucial to place them in the right environment. Expand
Deep Pyramidal Residual Networks
TLDR
This research gradually increases the feature map dimension at all units to involve as many locations as possible in the network architecture and proposes a novel residual unit capable of further improving the classification accuracy with the new network architecture. Expand
...
1
2
3
4
5
...