Corpus ID: 6530726

Challenges in Disentangling Independent Factors of Variation

@article{Szab2018ChallengesID,
  title={Challenges in Disentangling Independent Factors of Variation},
  author={A. Szab{\'o} and Qiyang Hu and Tiziano Portenier and Matthias Zwicker and P. Favaro},
  journal={ArXiv},
  year={2018},
  volume={abs/1711.02245}
}
  • A. Szabó, Qiyang Hu, +2 authors P. Favaro
  • Published 2018
  • Computer Science
  • ArXiv
  • We study the problem of building models that disentangle independent factors of variation. Such models could be used to encode features that can efficiently be used for classification and to transfer attributes between different images in image synthesis. As data we use a weakly labeled training set. Our weak labels indicate what single factor has changed between two data samples, although the relative value of the change is unknown. This labeling is of particular interest as it may be readily… CONTINUE READING
    37 Citations

    Figures, Tables, and Topics from this paper

    Disentangling Factors of Variation with Cycle-Consistent Variational Auto-Encoders
    • 40
    • Highly Influenced
    • PDF
    DualDis: Dual-Branch Disentangling with Adversarial Learning
    • 1
    • PDF
    Disentangling Factors of Variation by Mixing Them
    • 39
    • PDF
    Disentangling Factors of Variation by Mixing Them Qiyang
    • 1
    Unsupervised Robust Disentangling of Latent Characteristics for Image Synthesis
    • 17
    • PDF
    Demystifying Inter-Class Disentanglement
    • 14
    • PDF
    Gated Variational AutoEncoders: Incorporating Weak Supervision to Encourage Disentanglement
    • 3
    • PDF
    PuppetGAN: Transferring Disentangled Properties from Synthetic to Real Images
    • 1
    Latent Optimization for Non-adversarial Representation Disentanglement
    • 2

    References

    SHOWING 1-10 OF 24 REFERENCES
    Disentangling factors of variation in deep representation using adversarial training
    • 319
    • Highly Influential
    • PDF
    The Variational Fair Autoencoder
    • 289
    • PDF
    Learning to Disentangle Factors of Variation with Manifold Interaction
    • 189
    • PDF
    InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
    • 2,255
    • PDF
    Weakly-supervised Disentangling with Recurrent Transformations for 3D View Synthesis
    • 249
    • PDF
    Generative Adversarial Nets
    • 20,163
    • PDF
    Discovering Hidden Factors of Variation in Deep Networks
    • 145
    • PDF
    Disentangled Representation Learning GAN for Pose-Invariant Face Recognition
    • L. Tran, Xi Yin, X. Liu
    • Computer Science
    • 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
    • 2017
    • 490
    • PDF
    Adversarial Feature Learning
    • 995
    • PDF
    Reconstruction-Based Disentanglement for Pose-Invariant Face Recognition
    • 79
    • PDF