paper-with-me

홈 › Papers

How to Not Measure Disentanglement

2019-10-12 · Anna Sepliarskaia, Julia Kiseleva, Maarten de Rijke

To evaluate disentangled representations several metrics have been proposed. However, theoretical guarantees for conventional metrics of disentanglement are missing. Moreover, conventional metrics do not have a consistent correlation with the outcomes of qualitative studies. In this paper we analyze metrics of disentanglement and their properties. We conclude that existing metrics of disentanglement were created to reflect different characteristics of disentanglement and do not satisfy two basic desirable properties: (1) assign a high score to representations that are disentangled according to the definition; and (2) assign a low score to representations that are entangled according to the definition. In addition, we propose a new metric of disentanglement and prove that it satisfies both of the properties.

📄 PDF Abstract BibTeX arXiv:1910.05587

Code (0)

등록된 구현이 없습니다.

Tasks

DisentanglementReinforcement LearningTransfer LearningZero-Shot Learning

Similar Papers 제목 키워드 기반

Defining and Measuring Disentanglement for non-Independent Factors of Variation

2024-08-13 · Antonio Almudévar, Alfonso Ortega, Luis Vicente, Antonio Miguel 외

Representation learning is an approach that allows to discover and extract the factors of variation from the data. Intuitively, a representation is said to be disentangled if it separates the different factors of variati…

DisentanglementRepresentation Learningvalid

Measuring the Biases and Effectiveness of Content-Style Disentanglement

2020-08-27 · Xiao Liu, Spyridon Thermos, Gabriele Valvano, Agisilaos Chartsias 외

A recent spate of state-of-the-art semi- and un-supervised solutions disentangle and encode image "content" into a spatial tensor and image appearance or "style" into a vector, to achieve good performance in spatially eq…

DisentanglementImage-to-Image Translation

Measuring the Effect of Causal Disentanglement on the Adversarial Robustness of Neural Network Models

2023-08-21 · Preben M. Ness, Dusica Marijan, Sunanda Bose

Causal Neural Network models have shown high levels of robustness to adversarial attacks as well as an increased capacity for generalisation tasks such as few-shot learning and rare-context classification compared to tra…

Adversarial RobustnessBenchmarkingDisentanglementFew-Shot Learning+2

Measuring Disentanglement: A Review of Metrics

2020-12-16 · Marc-André Carbonneau, Julian Zaidi, Jonathan Boilard, Ghyslain Gagnon

Learning to disentangle and represent factors of variation in data is an important problem in AI. While many advances have been made to learn these representations, it is still unclear how to quantify disentanglement. Wh…

Disentanglement

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

2025-02-05 · Xingshen Zhang, Lin Wang, Shuangrong Liu, Xintao Lu 외

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 rep…

DisentanglementRepresentation Learning