COMPLETER: Incomplete Multi-view Clustering via Contrastive Prediction

Lin, Yijie; Gou, Yuanbiao; Liu, Zitao; Li, Boyun; Lv, Jiancheng; Peng, Xi

doi:10.1109/cvpr46437.2021.01102

Cited by 184 publications

(86 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, multi-view clustering were discussed with more techniques, e.g., binary coding [51] and self-paced learning [36]. Deep model based multi-view clustering [24,40,45,46,54] also attracted increasing attention in recent years.…”

Section: Related Workmentioning

confidence: 99%

Multi-VAE: Learning Disentangled View-common and View-peculiar Visual Representations for Multi-view Clustering

Xu¹,

Ren²,

Tang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Multi-view clustering, a long-standing and important research problem, focuses on mining complementary information from diverse views. However, existing works often fuse multiple views' representations or handle clustering in a common feature space, which may result in their entanglement especially for visual representations. To address this issue, we present a novel VAE-based multi-view clustering framework (Multi-VAE) by learning disentangled visual representations. Concretely, we define a view-common variable and multiple view-peculiar variables in the generative model. The prior of view-common variable obeys approximately discrete Gumbel Softmax distribution, which is introduced to extract the common cluster factor of multiple views. Meanwhile, the prior of view-peculiar variable follows continuous Gaussian distribution, which is used to represent each view's peculiar visual factors. By controlling the mutual information capacity to disentangle the viewcommon and view-peculiar representations, continuous visual information of multiple views can be separated so that their common discrete cluster information can be effectively mined. Experimental results demonstrate that Multi-VAE enjoys the disentangled and explainable visual representations, while obtaining superior clustering performance compared with state-of-the-art methods.

show abstract

Section: Related Workmentioning

confidence: 99%

Multi-VAE: Learning Disentangled View-common and View-peculiar Visual Representations for Multi-view Clustering

Xu¹,

Ren²,

Tang³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Deep Incomplete Multi-view Clustering. With the development of deep learning, a number of recent studies (Zhang et al 2020;Lin et al 2021) employ DNNs to solve incomplete multi-view clustering challenges, showing noticeable improvement on clustering performance. Based on producing missing data or not, most existing deep incomplete multi-view clustering methods could be roughly classified into two categories, i.e., learning a uniform representation without generating missing data (Wen et al 2020b,a;Wang et al 2021b) and generating missing instances using adversarial learning (Wang et al 2021a(Wang et al , 2018Zhang et al 2020;Xu et al 2021).…”

Section: Related Workmentioning

confidence: 99%

“…GP-MVC (Wang et al 2021a) introduces view-specific GAN with multi-view cycle consistency to generate the missing data of one view conditioning on the shared representation given by other views. Unlike these two types of methods, COMPLETER (Lin et al 2021) generates the representations of the missing views by minimizing the conditional entropy of different views by dual prediction. Different from the current methods, we focus on generating missing data based on the similarity relationships of corresponding instances in the existing views.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Incomplete Multi-view Clustering via Cross-view Relation Transfer

Wang¹,

Chang²,

Fu³

et al. 2021

Preprint

View full text Add to dashboard Cite

In this paper, we consider the problem of multi-view clustering on incomplete views. Compared with complete multiview clustering, the view-missing problem increases the difficulty of learning common representations from different views. To address the challenge, we propose a novel incomplete multi-view clustering framework, which incorporates cross-view relation transfer and multi-view fusion learning. Specifically, based on the consistency existing in multi-view data, we devise a cross-view relation transfer-based completion module, which transfers known similar inter-instance relationships to the missing view and recovers the missing data via graph networks based on the transferred relationship graph. Then the view-specific encoders are designed to extract the recovered multi-view data, and an attention-based fusion layer is introduced to obtain the common representation. Moreover, to reduce the impact of the error caused by the inconsistency between views and obtain a better clustering structure, a joint clustering layer is introduced to optimize recovery and clustering simultaneously. Extensive experiments conducted on several real datasets demonstrate the effectiveness of the proposed method.

show abstract

“…To be able to explain learned representations would provide crucial information in several use-cases. For instance, a typical clustering approach is applying K-means to the representation produced by a feature extractor trained on unlabeled data [28,50,55], but there is no method for investigating which features are characteristic for the members of a cluster. Representation learning explainability would also allow for a new approach for evaluating representation learning frameworks.…”

Section: Introductionmentioning

confidence: 99%

RELAX: Representation Learning Explainability

Wickstrøm¹,

Trosten²,

Løkse³

et al. 2021

Preprint

View full text Add to dashboard Cite

Despite the significant improvements that representation learning via self-supervision has led to when learning from unlabeled data, no methods exist that explain what influences the learned representation. We address this need through our proposed approach, RELAX, which is the first approach for attribution-based explanations of representations. Our approach can also model the uncertainty in its explanations, which is essential to produce trustworthy explanations. RELAX explains representations by measuring similarities in the representation space between an input and masked out versions of itself, providing intuitive explanations and significantly outperforming the gradient-based baseline. We provide theoretical interpretations of RELAX and conduct a novel analysis of feature extractors trained using supervised and unsupervised learning, providing insights into different learning strategies. Finally, we illustrate the usability of RELAX in multi-view clustering and highlight that incorporating uncertainty can be essential for providing low-complexity explanations, taking a crucial step towards explaining representations.

show abstract

COMPLETER: Incomplete Multi-view Clustering via Contrastive Prediction

Cited by 184 publications

References 28 publications

Multi-VAE: Learning Disentangled View-common and View-peculiar Visual Representations for Multi-view Clustering

Multi-VAE: Learning Disentangled View-common and View-peculiar Visual Representations for Multi-view Clustering

Incomplete Multi-view Clustering via Cross-view Relation Transfer

RELAX: Representation Learning Explainability

Contact Info

Product

Resources

About