mlsanityData to 7 Oct 2026

DualContrast: Unsupervised Disentangling of Content and Transformations with Implicit Parameterization

arXiv 2405.16796 ↗First posted 27 May 2024cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Unsupervised disentanglement of content and transformation is significantly important for analyzing shape-focused scientific image datasets, given their efficacy in solving downstream image-based shape-analyses tasks. The existing relevant works address the problem by explicitly parameterizing the transformation latent codes in a generative model, significantly reducing their expressiveness. Moreover, they are not applicable in cases where transformations can not be readily parametrized. An alternative to such explicit approaches is contrastive methods with data augmentation, which implicitly disentangles transformations and content. However, the existing contrastive strategies are insufficient to this end. Therefore, we developed a novel contrastive method with generative modeling, DualContrast, specifically for unsupervised disentanglement of content and transformations in shape-focused image datasets. DualContrast creates positive and negative pairs for content and transformation from data and latent spaces. Our extensive experiments showcase the efficacy of DualContrast over existing self-supervised and explicit parameterization approaches. With DualContrast, we disentangled protein composition and conformations in cellular 3D protein images, which was unattainable with existing disentanglement approaches

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 34 of its 55 references that are in the map. First posted on arXiv 27 May 2024.

Builds on

34 of 55 references.

References · 34

  1. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  2. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  3. 2015Spatial Transformer Networkscited by 1,423
  4. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,186
  5. 2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 929
  6. 2016InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Netscited by 858
  7. 2018Disentangling by Factorisingcited by 614
  8. 2016Geometric deep learning on graphs and manifolds using mixture model CNNscited by 468
  9. 2021Self-Supervised Learning with Data Augmentations Provably Isolates Content from Stylecited by 308
  10. 2017Variational Inference of Disentangled Latent Concepts from Unlabeled Observationscited by 240
  11. 2023StyleDiffusion: Controllable Disentangled Style Transfer via Diffusion Modelscited by 154
  12. 2021Learning Debiased Representation via Disentangled Feature Augmentationcited by 116
22 more
  1. 2021NPMs: Neural Parametric Models for 3D Deformable Shapescited by 89
  2. 2016Disentangling factors of variation in deep representations using adversarial trainingcited by 87
  3. 2017Variational Autoencoders for Deforming 3D Mesh Modelscited by 69
  4. 2020Unsupervised Shape and Pose Disentanglement for 3D Meshescited by 61
  5. 2015Learning Descriptors for Object Recognition and 3D Pose Estimationcited by 60
  6. 2020LIMP: Learning Latent Shape Representations with Metric Preservation Priorscited by 54
  7. 2018Disentangling Factors of Variation with Cycle-Consistent Variational Auto-Encoderscited by 49
  8. 2021Learning Disentangled Representations in the Imaging Domaincited by 46
  9. 2021ARAPReg: An As-Rigid-As Possible Regularization Loss for Learning Deformable Shape Generatorscited by 40
  10. 2021Diagonal Attention and Style-based GAN for Content-Style Disentanglement in Image Generation and Translationcited by 35
  11. 2021Learning Disentangled Representation by Exploiting Pretrained Generative Models: A Contrastive Learning Viewcited by 34
  12. 2019Geometric Disentanglement for Generative Latent Shape Modelscited by 30
  13. 2021Unsupervised Part Segmentation through Disentangling Appearance and Shapecited by 29
  14. 2019Explicitly disentangling image content from translation and rotation with spatial-VAEcited by 28
  15. 2022CryoAI: Amortized Inference of Poses for Ab Initio Reconstruction of 3D Molecular Volumes from Real Cryo-EM Imagescited by 24
  16. 2022Amortized Inference for Heterogeneous Reconstruction in Cryo-EMcited by 23
  17. 2019Explicit Disentanglement of Appearance and Perspective in Generative Modelscited by 20
  18. 2022A Contrastive Objective for Learning Disentangled Representationscited by 14
  19. 2023Simple Disentanglement of Style and Content in Visual Representationscited by 10
  20. 2021Rethinking Content and Style: Exploring Bias for Unsupervised Disentanglementcited by 7
  21. 2019Bayes-Factor-VAE: Hierarchical Bayesian Deep Auto-Encoder Models for Factor Disentanglementcited by 6
  22. 2019A complete data processing workflow for CryoET and subtomogram averagingcited by 5

Built on by

Not cited yet.