Pairwise Similarity Regularization for Adversarial Domain Adaptation

Wang, Haotian; Yang, Wenjing; Wang, Ji; Wang, Ruxin; Lan, Long; Geng, Mingyang

doi:10.1145/3394171.3413516

Cited by 10 publications

(6 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such models are most commonly used for image recognition and classification purposes. Thereby, they usually contain some FC layers at the very end of the model (Wang et al, 2020) and are often referred to as encoders. But we do not want our model to produce just one or a vector of values for a given input.…”

Section: Choice Of Model Architecturementioning

confidence: 99%

A deep learning solution for crystallographic structure determination

Pan

Jin

Miller

et al. 2023

IUCrJ

View full text Add to dashboard Cite

The general de novo solution of the crystallographic phase problem is difficult and only possible under certain conditions. This paper develops an initial pathway to a deep learning neural network approach for the phase problem in protein crystallography, based on a synthetic dataset of small fragments derived from a large well curated subset of solved structures in the Protein Data Bank (PDB). In particular, electron-density estimates of simple artificial systems are produced directly from corresponding Patterson maps using a convolutional neural network architecture as a proof of concept.

show abstract

Section: Choice Of Model Architecturementioning

confidence: 99%

A deep learning solution for crystallographic structure determination

Pan

Jin

Miller

et al. 2023

IUCrJ

View full text Add to dashboard Cite

show abstract

“…Wang et al. [9] introduce a pairwise similarity regularisation method based on adversarial learning, which exploits the clustering structure of the target domain and minimises the difference between the clustered partitions and the pseudo‐predicted pairwise similarities. Wang et al.…”

Section: Related Workmentioning

confidence: 99%

“…Later, Long et al [8] come up with a principle framework conditioned on setting an adversarial adaptation model based on the discriminatory information conveyed in the classifier predictions. Wang et al [9] introduce a pairwise similarity regularisation method based on adversarial learning, which exploits the clustering structure of the target domain and minimises the difference between the clustered partitions and the pseudo-predicted pairwise similarities. Wang et al [4] propose an interchangeable batch normalisation, a novel module for UDA, to effectively fuse different channels for UDA.…”

Section: Vanilla Unsupervised Domain Adaptationmentioning

confidence: 99%

See 1 more Smart Citation

Domain‐specific feature recalibration and alignment for multi‐source unsupervised domain adaptation

Wang

Chen

Tan

et al. 2022

IET Computer Vision

Self Cite

View full text Add to dashboard Cite

Traditional unsupervised domain adaptation (UDA) usually assumes that the source domain has labels and the target domain has no labels. In a real environment, labelled source domain data usually comes from multiple different distributions. To handle this problem, multi‐source unsupervised domain adaptation (MUDA) is proposed. Multi‐source unsupervised domain adaptation aims to adapt the model trained on multi‐labelled source domains to the unlabelled target domain. In this paper, a novel MUDA method by domain‐specific feature recalibration and alignment (FRA) is proposed. Specifically, to achieve feature recalibration, the authors leverage channel attention to pick out significant channels and spatial attention to focus on important features in different channels. Such integration of channel and spatial attention can lead to effective domain‐specific feature recalibration that may be of great importance to MUDA. In addition, to achieve better MUDA, the authors propose domain‐specific feature alignment which consists of Maximum Mean Discrepancy and JS‐divergence loss. Maximum Mean Discrepancy can reduce the difference between the source domain and target domain. Meanwhile, JS‐divergence loss may ensure the prediction consistency of different classifiers in the source domains. Four experiments have proved that FRA can achieve significantly better results in popular benchmarks for MUDA.

show abstract

“…To overcome the above problem, the research community has made great efforts to address domain shift. Specifically, unsupervised domain adaptation (UDA) receives a lot of attention [8,26,44] since it requires no label information in target domain. The latest work [10,47] based on deep learning has achieved remarkable results.…”

Section: Introductionmentioning

confidence: 99%

InterBN: Channel Fusion for Adversarial Unsupervised Domain Adaptation

Wang

Li³

et al. 2021

Proceedings of the 29th ACM International Conference on Multimedia

Self Cite

View full text Add to dashboard Cite

A classifier trained on one dataset rarely works on other datasets obtained under different conditions because of domain shifting. Such a problem is usually solved by domain adaptation methods. In this paper, we propose a novel unsupervised domain adaptation (UDA) method based on Interchangeable Batch Normalization (InterBN) to fuse different channels in deep neural networks for adversarial domain adaptation. Specifically, we first observe that the channels with small batch normalization scaling factor have less influence on the whole domain adaption, followed by a theoretical proof that the scaling factors for some channels will definitely come close to zero when imposing a sparsity regularization. Then, we replace the channels that have smaller scaling factors in the source domain with the mean of the channels which have larger scaling factors in the target domain or vice versa. Such a simple but effective channel fusion scheme can drastically increase the domain adaption ability. Extensive experimental results show that our InterBN significantly outperforms the current adversarial domain adaptation methods by a large margin on four visual benchmarks. In particular, InterBN achieves a remarkable improvement of 7.7% over the conditional adversarial adaptation networks (CDAN) on VisDA-2017 benchmark. CCS CONCEPTS• Networks → Network architectures; Network design principles.

show abstract

Pairwise Similarity Regularization for Adversarial Domain Adaptation

Cited by 10 publications

References 31 publications

A deep learning solution for crystallographic structure determination

A deep learning solution for crystallographic structure determination

Domain‐specific feature recalibration and alignment for multi‐source unsupervised domain adaptation

InterBN: Channel Fusion for Adversarial Unsupervised Domain Adaptation

Contact Info

Product

Resources

About