Corpus ID: 236090250

Domain Generalization in Vision: A Survey

@inproceedings{Zhou2021DomainGI,
  title={Domain Generalization in Vision: A Survey},
  author={Kaiyang Zhou and Ziwei Liu and Yu Qiao and Tao Xiang and Chen Change Loy},
  year={2021}
}
Generalization to out-of-distribution (OOD) data is a capability natural to humans yet challenging for machines to reproduce. This is because most learning algorithms strongly rely on the i.i.d. assumption on source/target data, which is often violated in practice due to domain shift. Domain generalization (DG) aims to achieve OOD generalization by using only source data for model learning. Since first introduced in 2011, research in DG has made great progresses. In particular, intensive… Expand

Figures and Tables from this paper

Energy-Based Open-World Uncertainty Modeling for Confidence Calibration
TLDR
This work proposes a novel K+1-way softmax formulation, which incorporates the modeling of open-world uncertainty as the extra dimension, and theoretically proves that optimizing such an objective essentially forces theextra dimension to capture the marginal data distribution. Expand
Detecting Mitosis against Domain Shift using a Fused Detector and Deep Ensemble Classification Model for MIDOG Challenge
TLDR
This work proposes a two stage mitotic figure detection framework by fusing a detector and a deep ensemble classification model, and utilizes both stain normalization and data augmentation, aiding model to learn color irrelevant features. Expand
Learning to Prompt for Vision-Language Models
TLDR
Experiments show that CoOp effectively turns pre-trained vision-language models into data-efficient visual learners, requiring as few as one or two shots to beat hand-crafted prompts with a decent margin and able to gain significant improvements when using more shots (e.g., at 16 shots the average gain is around 17% with the highest reaching over 50%). Expand

References

SHOWING 1-10 OF 179 REFERENCES
Robust Place Categorization With Deep Domain Generalization
TLDR
This paper proposes a novel deep learning framework for domain generalization that exploits recent advances in deep domain adaptation and design a convolutional neural network architecture with novel layers performing a weighted version of batch normalization. Expand
Best Sources Forward: Domain Generalization through Source-Specific Nets
TLDR
This work designs a deep network with multiple domain-specific classifiers, each associated to a source domain, and introduced a domain agnostic component supporting the final classifier. Expand
Adapting Visual Category Models to New Domains
TLDR
This paper introduces a method that adapts object models acquired in a particular visual domain to new imaging conditions by learning a transformation that minimizes the effect of domain-induced changes in the feature distribution. Expand
Zero-Shot Domain Generalization
TLDR
This work proposes a simple strategy which effectively exploits semantic information of classes, to adapt existing DG methods to meet the demands of Zero-Shot Domain Generalization, and evaluates the proposed methods on CIFAR-10, CIFar-100, F-MNIST and PACS datasets. Expand
Deeper, Broader and Artier Domain Generalization
TLDR
This paper builds upon the favorable domain shift-robust properties of deep learning methods, and develops a low-rank parameterized CNN model for end-to-end DG learning that outperforms existing DG alternatives. Expand
Learning Attributes Equals Multi-Source Domain Generalization
TLDR
This work provides a novel perspective to attribute detection and proposes to gear the techniques in multi-source domain generalization for the purpose of learning cross-category generalizable attribute detectors. Expand
Multi-view Domain Generalization for Visual Recognition
  • Li Niu, Wen Li, Dong Xu
  • Mathematics, Computer Science
  • 2015 IEEE International Conference on Computer Vision (ICCV)
  • 2015
TLDR
This paper builds upon exemplar SVMs to learn a set of SVM classifiers by using one positive sample and all negative samples in the source domain each time, and introduces a new regularizer to minimize the mismatch between any two representation matrices on different views. Expand
Domain Generalization for Object Recognition with Multi-task Autoencoders
TLDR
This work proposes a new feature learning algorithm, Multi-Task Autoencoder (MTAE), that provides good generalization performance for cross-domain object recognition and evaluates the performance of the algorithm on benchmark image recognition datasets, where the task is to learn features from multiple datasets and to then predict the image label from unseen datasets. Expand
Automated Synthetic-to-Real Generalization
TLDR
This work explicitly encourages the synthetically trained model to maintain similar representations with the ImageNet pre-trained model, and proposes a learning-to-optimize (L2O) strategy to automate the selection of layer-wise learning rates. Expand
Towards Recognizing Unseen Categories in Unseen Domains
TLDR
CuMix (CurriculumMixup for recognizing unseen categories in unseen domains), a holistic algorithm to tackle ZSL, DG and ZSL+DG, to simulate the test-time domain and semantic shift using images and features from unseen domains and categories generated by mixing up the multiple source domains and Categories available during training. Expand
...
1
2
3
4
5
...