paper-with-me

Papers

Disentanglement in Difference: Directly Learning Semantically Disentangled Representations by Maximizing Inter-Factor Differences

2025-02-05 · Xingshen Zhang, Lin Wang, Shuangrong Liu, Xintao Lu, Chaoran Pang, Bo Yang

In this study, Disentanglement in Difference(DiD) is proposed to address the inherent inconsistency between the statistical independence of latent variables and the goal of semantic disentanglement in disentanglement representation learning. Conventional disentanglement methods achieve disentanglement representation by improving statistical independence among latent variables. However, the statistical independence of latent variables does not necessarily imply that they are semantically unrelated, thus, improving statistical independence does not always enhance disentanglement performance. To address the above issue, DiD is proposed to directly learn semantic differences rather than the statistical independence of latent variables. In the DiD, a Difference Encoder is designed to measure the semantic differences; a contrastive loss function is established to facilitate inter-dimensional comparison. Both of them allow the model to directly differentiate and disentangle distinct semantic factors, thereby resolving the inconsistency between statistical independence and semantic disentanglement. Experimental results on the dSprites and 3DShapes datasets demonstrate that the proposed DiD outperforms existing mainstream methods across various disentanglement metrics.

📄 PDF Abstract BibTeX arXiv:2502.03123

Code (0)

등록된 구현이 없습니다.

Tasks

DisentanglementRepresentation Learning

Similar Papers 제목 키워드 기반

Variational Autoencoder with Disentanglement Priors for Low-Resource Task-Specific Natural Language Generation

2022-02-27 · Zhuang Li, Lizhen Qu, Qiongkai Xu, Tongtong Wu 외

In this paper, we propose a variational autoencoder with disentanglement priors, VAE-DPRIOR, for task-specific natural language generation with none or a handful of task-specific labeled examples. In order to tackle comp…

Data AugmentationDisentanglementFew-Shot LearningRepresentation Learning+3

Learning Disentangled Representation by Exploiting Pretrained Generative Models: A Contrastive Learning View

2021-02-21 · ICLR 2022 4 · Xuanchi Ren, Tao Yang, Yuwang Wang, Wenjun Zeng

From the intuitive notion of disentanglement, the image variations corresponding to different factors should be distinct from each other, and the disentangled representation should reflect those variations with separate …

Contrastive LearningDisentanglementRepresentation Learning

Representation Disentanglement for Multi-task Learning with application to Fetal Ultrasound

2019-08-21 · Qingjie Meng, Nick Pawlowski, Daniel Rueckert, Bernhard Kainz

One of the biggest challenges for deep learning algorithms in medical image analysis is the indiscriminate mixing of image properties, e.g. artifacts and anatomy. These entangled image properties lead to a semantically r…

AnatomyDisentanglementMedical Image AnalysisMulti-Task Learning

Physically Disentangled Representations

2022-04-11 · Tzofi Klinghoffer, Kushagra Tiwary, Arkadiusz Balata, Vivek Sharma 외

State-of-the-art methods in generative representation learning yield semantic disentanglement, but typically do not consider physical scene parameters, such as geometry, albedo, lighting, or camera. We posit that inverse…

AttributeClassificationDisentanglementEmotion Recognition+2

Learning Disentangled Representations of Negation and Uncertainty

2022-04-01 · ACL 2022 5 · Jake Vasilakes, Chrysoula Zerva, Makoto Miwa, Sophia Ananiadou

Negation and uncertainty modeling are long-standing tasks in natural language processing. Linguistic theory postulates that expressions of negation and uncertainty are semantically independent from each other and the con…

DisentanglementNegationRepresentation Learning