Movatterモバイル変換


[0]ホーム

URL:


Skip to main content

Advertisement

Springer Nature Link
Log in

A Simple Method to Evaluate Support Size and Non-uniformity of a Decoder-Based Generative Model

  • Conference paper
  • First Online:

Part of the book series:Lecture Notes in Computer Science ((LNISA,volume 11832))

  • 1132Accesses

Abstract

Theoretical analysis in [1] suggested that adversarially trained generative models are naturally inclined to learn distribution with low support. In particular, this effect is caused by the limited capacity of the discriminator network. To verify this claim, [2] proposed a statistical test based on the birthday paradox that partially confirmed the analysis. In this paper, we continue this line of work and develop a parameter-free and straightforward method to estimate the support size of an arbitrary decoder-based generative model. Our approach considers the decoder network from a geometric viewpoint and evaluates the support size as the volume of the manifold containing the generative model samples. Additionally, we propose a method to measure non-uniformity of a generative model that can provide additional insight into the model’s behavior. We then apply these tools to perform a quantitative comparison of common generative models.

This is a preview of subscription content,log in via an institution to check access.

Access this chapter

Subscribe and save

Springer+ Basic
¥17,985 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Chapter
JPY 3498
Price includes VAT (Japan)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
JPY 5719
Price includes VAT (Japan)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
JPY 7149
Price includes VAT (Japan)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide -see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Similar content being viewed by others

Notes

  1. 1.

    \(ReLU(x) = \max (0, x)\).

References

  1. Arora, S., Ge, R., Liang, Y., Ma, T., Zhang, Y.: Generalization and equilibrium in generative adversarial nets (GANs). arXiv preprintarXiv:1703.00573 (2017)

  2. Arora, S., Zhang, Y.: Do GANs actually learn the distribution? An empirical study. arXiv preprintarXiv:1706.08224 (2017)

  3. Ben-Israel, A.: The change-of-variables formula using matrix volume. SIAM J. Matrix Anal. Appl.21(1), 300–312 (1999)

    Article MathSciNet  Google Scholar 

  4. Chen, X., Duan, Y., Houthooft, R., Schulman, J., Sutskever, I., Abbeel, P.: InfoGAN: interpretable representation learning by information maximizing generative adversarial nets. In: Advances in Neural Information Processing Systems, pp. 2172–2180 (2016)

    Google Scholar 

  5. Dinh, L., Sohl-Dickstein, J., Bengio, S.: Density estimation using real NVP. arXiv preprintarXiv:1605.08803 (2016)

  6. Goodfellow, I., et al.: Generative adversarial nets. In: Advances in Neural Information Processing Systems, pp. 2672–2680 (2014)

    Google Scholar 

  7. Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., Klambauer, G., Hochreiter, S.: GANs trained by a two time-scale update rule converge to a nash equilibrium. arXiv preprintarXiv:1706.08500 (2017)

  8. Kingma, D.P., Salimans, T., Jozefowicz, R., Chen, X., Sutskever, I., Welling, M.: Improved variational inference with inverse autoregressive flow. In: Advances in Neural Information Processing Systems, pp. 4743–4751 (2016)

    Google Scholar 

  9. Kingma, D.P., Welling, M.: Auto-encoding variational Bayes. arXiv preprintarXiv:1312.6114 (2013)

  10. Lucic, M., Kurach, K., Michalski, M., Gelly, S., Bousquet, O.: Are GANs created equal. A large-scale study. arXiv e-prints (2017)

    Google Scholar 

  11. Metz, L., Poole, B., Pfau, D., Sohl-Dickstein, J.: Unrolled generative adversarial networks. arXiv preprintarXiv:1611.02163 (2016)

  12. Odena, A., et al.: Is generator conditioning causally related to GAN performance? arXiv preprintarXiv:1802.08768 (2018)

  13. Oord, A.V.D., Kalchbrenner, N., Kavukcuoglu, K.: Pixel recurrent neural networks. arXiv preprintarXiv:1601.06759 (2016)

  14. Sajjadi, M.S., Bachem, O., Lucic, M., Bousquet, O., Gelly, S.: Assessing generative models via precision and recall. arXiv preprintarXiv:1806.00035 (2018)

  15. Salakhutdinov, R., Larochelle, H.: Efficient learning of deep Boltzmann machines. In: Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, pp. 693–700 (2010)

    Google Scholar 

  16. Salimans, T., Goodfellow, I., Zaremba, W., Cheung, V., Radford, A., Chen, X.: Improved techniques for training GANs. In: Advances in Neural Information Processing Systems, pp. 2234–2242 (2016)

    Google Scholar 

  17. Spivak, M.: Calculus on Manifolds. W.A. Benjamin Inc., Massachusetts (1965)

    MATH  Google Scholar 

  18. Theis, L., Oord, A.V.D., Bethge, M.: A note on the evaluation of generative models. arXiv preprintarXiv:1511.01844 (2015)

  19. Wu, Y., Burda, Y., Salakhutdinov, R., Grosse, R.: On the quantitative analysis of decoder-based generative models. arXiv preprintarXiv:1611.04273 (2016)

Download references

Acknowledgements

The authors would like to thank Ilya Tolstikhin for fruitful discussions. The work was partly supported by Ministry of Education and Science of the Russian Federation (grant 14.756.31.0001). Dmitry Vetrov was also partly supported by Samsung Research, Samsung Electronics.

Author information

Authors and Affiliations

  1. National Research University Higher School of Economics, Moscow, Russia

    Kirill Struminsky

  2. Skolkovo Institute of Science and Technology, Moscow, Russia

    Kirill Struminsky

  3. Samsung-HSE Laboratory, National Research University Higher School of Economics, Moscow, Russia

    Dmitry Vetrov

Authors
  1. Kirill Struminsky

    You can also search for this author inPubMed Google Scholar

  2. Dmitry Vetrov

    You can also search for this author inPubMed Google Scholar

Corresponding author

Correspondence toKirill Struminsky.

Editor information

Editors and Affiliations

  1. RWTH Aachen University, Aachen, Germany

    Wil M. P. van der Aalst

  2. University of Ljubljana, Ljubljana, Slovenia

    Vladimir Batagelj

  3. National Research University Higher School of Economics, Moscow, Russia

    Dmitry I. Ignatov

  4. Krasovskii Institute of Mathematics and Mechanics, Yekaterinburg, Russia

    Michael Khachay

  5. National Research University Higher School of Economics, Moscow, Russia

    Valentina Kuskova

  6. University of Oslo, Oslo, Norway

    Andrey Kutuzov

  7. National Research University Higher School of Economics, Moscow, Russia

    Sergei O. Kuznetsov

  8. National Research University Higher School of Economics, Moscow, Russia

    Irina A. Lomazova

  9. Lomonosov Moscow State University, Moscow, Russia

    Natalia Loukachevitch

  10. LORIA, Vandœuvre-lès-Nancy, France

    Amedeo Napoli

  11. University of Florida, Gainesville, FL, USA

    Panos M. Pardalos

  12. Ca Foscari University of Venice, Venice, Italy

    Marcello Pelillo

  13. National Research University Higher School of Economics, Nizhny Novgorod, Russia

    Andrey V. Savchenko

  14. Kazan Federal University, Kazan, Russia

    Elena Tutubalina

Rights and permissions

Copyright information

© 2019 Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Struminsky, K., Vetrov, D. (2019). A Simple Method to Evaluate Support Size and Non-uniformity of a Decoder-Based Generative Model. In: van der Aalst, W.,et al. Analysis of Images, Social Networks and Texts. AIST 2019. Lecture Notes in Computer Science(), vol 11832. Springer, Cham. https://doi.org/10.1007/978-3-030-37334-4_8

Download citation

Publish with us

Access this chapter

Subscribe and save

Springer+ Basic
¥17,985 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Chapter
JPY 3498
Price includes VAT (Japan)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
JPY 5719
Price includes VAT (Japan)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
JPY 7149
Price includes VAT (Japan)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide -see info

Tax calculation will be finalised at checkout

Purchases are for personal use only


[8]ページ先頭

©2009-2025 Movatter.jp