Detach and Adapt: Learning Cross-Domain Disentangled Deep Representation

Liu, Yen‐Cheng; Yeh, Yu-Ying; Fu, Tzu-Chien; Wang, Sheng‐De; Chiu, Wei-Chen; Wang, Yu-Chiang Frank

doi:10.48550/arxiv.1705.01314

Cited by 4 publications

(5 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The third family performs both marginal-distribution-difference elimination and syntheticdata generation, such as Cross-Domain Representation Disentangler (CDRD) [15], Synthesized Examples for Generalized Zero-Shot Learning (SE-GZSL) [30], Disentangled Synthesis for Domain Adaptation (DiDA) [6], Attribute-Based Synthetic Network (ABS-Net) [18], among others. Madras et al [20] proposed such a FML framework.…”

Section: Hybrid Methodsmentioning

confidence: 99%

“…For the second thrust that generates data with unseen class, domain combinations, we chose ELEGANT [35] which only uses domain labels and ML-VAE [5] which only uses class labels. For the third thrust that uses hybrid solutions, we chose ABS-Net [18] which is the base method of ours without an adversarial mechanism, and CDRD [15] and SE-GZSL [30], which can be treated as advanced instantiated algorithms under the FML framework of Madras et al [20]. Finally, we compare the direct learning strategy that stacks P, G 1 , and D 11 as the whole network.…”

Section: Methods For Comparisonmentioning

confidence: 99%

See 1 more Smart Citation

Additive Adversarial Learning for Unbiased Authentication

Liang

Cao

Zhang

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

Authentication is a task aiming to confirm the truth between data instances and personal identities. Typical authentication applications include face recognition, person re-identification, authentication based on mobile devices and so on. The recently-emerging data-driven authentication process may encounter undesired biases, i.e., the models are often trained in one domain (e.g., for people wearing spring outfits) while required to apply in other domains (e.g., they change the clothes to summer outfits). To address this issue, we propose a novel two-stage method that disentangles the class/identity from domaindifferences, and we consider multiple types of domaindifference. In the first stage, we learn disentangled representations by a one-versus-rest disentangle learning (OVRDL) mechanism. In the second stage, we improve the disentanglement by an additive adversarial learning (AAL) mechanism. Moreover, we discuss the necessity to avoid a learning dilemma due to disentangling causally related types of domain-difference. Comprehensive evaluation results demonstrate the effectiveness and superiority of the proposed method.

show abstract

Section: Hybrid Methodsmentioning

confidence: 99%

Section: Methods For Comparisonmentioning

confidence: 99%

Additive Adversarial Learning for Unbiased Authentication

Liang

Cao

Zhang

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

show abstract

“…This provides a considerable improvement for some cross-domain recognition tasks [37,23,34,26,30,21,6,9]. Specifically, a number of deep domain adaptation models have applied the adversarial training strategy [35,36,8,21,5,20,22]. DANN [8] employs a gradient reversal layer between the feature layer and the domain discriminator, causing feature representation to anti-learn the domain difference and hence adapt well to the target domain.…”

Section: Related Workmentioning

confidence: 99%

Unsupervised Domain Adaptation using Deep Networks with Cross-Grafted Stacks

Hou

Ding

Deng

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)

View full text Add to dashboard Cite

Current deep domain adaptation methods used in computer vision have mainly focused on learning discriminative and domain-invariant features across different domains. In this paper, we present a novel approach that bridges the domain gap by projecting the source and target domains into a common association space through an unsupervised "cross-grafted representation stacking" (CGRS) mechanism. Specifically, we construct variational auto-encoders (VAE) for the two domains, and form bidirectional associations by cross-grafting the VAEs' decoder stacks. Furthermore, generative adversarial networks (GAN) are employed for label alignment (LA), mapping the target domain data to the known label space of the source domain. The overall adaptation process hence consists of three phases: feature representation learning by VAEs, association generation, and association label alignment by GANs. Experimental results demonstrate that our CGRS-LA approach outperforms the state-of-the-art on a number of unsupervised domain adaptation benchmarks.

show abstract

“…Various proposals for more structure imposing regularization have been made, either with some sort of supervision (e.g. Siddharth et al, 2017;Bouchacourt et al, 2017;Liu et al, 2017;Mathieu et al, 2016;Cheung et al, 2014) or completely unsupervised (e.g. Higgins et al, 2017;Kim & Mnih, 2018;Chen et al, 2018;Kumar et al, 2018;Esmaeili et al, 2018).…”

Section: Related Workmentioning

confidence: 99%

Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustness

Suter¹,

Miladinović²,

Schölkopf³

et al. 2018

Preprint

View full text Add to dashboard Cite

The ability to learn disentangled representations that split underlying sources of variation in high dimensional, unstructured data is important for data efficient and robust use of neural networks. While various approaches aiming towards this goal have been proposed in recent times, a commonly accepted definition and validation procedure is missing. We provide a causal perspective on representation learning which covers disentanglement and domain shift robustness as special cases. Our causal framework allows us to introduce a new metric for the quantitative evaluation of deep latent variable models. We show how this metric can be estimated from labeled observational data and further provide an efficient estimation algorithm that scales linearly in the dataset size.

show abstract

Detach and Adapt: Learning Cross-Domain Disentangled Deep Representation

Cited by 4 publications

References 0 publications

Additive Adversarial Learning for Unbiased Authentication

Additive Adversarial Learning for Unbiased Authentication

Unsupervised Domain Adaptation using Deep Networks with Cross-Grafted Stacks

Robustly Disentangled Causal Mechanisms: Validating Deep Representations for Interventional Robustness

Contact Info

Product

Resources

About