Abstract
Generative models based on normalizing flows are very successful in modeling complex data distributions using simpler ones. However, straightforward linear interpolations show unexpected side effects, as interpolation paths lie outside the area where samples are observed. This is caused by the standard choice of Gaussian base distributions and can be seen in the norms of the interpolated samples as they are outside the data manifold. This observation suggests that changing the way of interpolating should generally result in better interpolations, but it is not clear how to do that in an unambiguous way. In this paper, we solve this issue by enforcing a specific manifold and, hence, change the base distribution, to allow for a principled way of interpolation. Specifically, we use the Dirichlet and von Mises-Fisher base distributions on the probability simplex and the hypersphere, respectively. Our experimental results show superior performance in terms of bits per dimension, Fréchet Inception Distance (FID), and Kernel Inception Distance (KID) scores for interpolation, while maintaining the generative performance.
S. G. Fadel and S. Mair—Equal contribution.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Agustsson, E., Sage, A., Timofte, R., Gool, L.V.: Optimal transport maps for distribution preserving operations on latent spaces of generative models. In: International Conference on Learning Representations (2019)
Arvanitidis, G., Hansen, L.K., Hauberg, S.: Latent space oddity: on the curvature of deep generative models. In: International Conference on Learning Representations (2018)
Bińkowski, M., Sutherland, D.J., Arbel, M., Gretton, A.: Demystifying MMD GANs. In: International Conference on Learning Representations (2018)
Bose, J., Smofsky, A., Liao, R., Panangaden, P., Hamilton, W.: Latent variable modelling with hyperbolic normalizing flows. In: International Conference on Machine Learning, pp. 1045–1055. PMLR (2020)
Brehmer, J., Cranmer, K.: Flows for simultaneous manifold learning and density estimation. In: Advances in Neural Information Processing Systems, vol. 33 (2020)
Chen, T.Q., Behrmann, J., Duvenaud, D.K., Jacobsen, J.H.: Residual flows for invertible generative modeling. In: Advances in Neural Information Processing Systems, pp. 9913–9923 (2019)
Clanuwat, T., Bober-Irizar, M., Kitamoto, A., Lamb, A., Yamamoto, K., Ha, D.: Deep learning for classical Japanese literature. CoRR abs/1812.01718 (2018)
Davidson, T.R., Falorsi, L., De Cao, N., Kipf, T., Tomczak, J.M.: Hyperspherical variational auto-encoders. In: Proceedings of the Thirty-Fourth Conference on Uncertainty in Artificial Intelligence, pp. 856–865 (2018)
Dinh, L., Krueger, D., Bengio, Y.: NICE: non-linear independent components estimation. In: 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, 7–9 May 2015, Workshop Track Proceedings (2015)
Dinh, L., Shol-Dickstein, J., Bengio, S.: Density estimation using Real NVP. In: International Conference on Learning Representations (2017)
Gemici, M.C., Rezende, D., Mohamed, S.: Normalizing flows on Riemannian manifolds. CoRR abs/1611.02304 (2016)
Goodfellow, I., et al.: Generative adversarial nets. In: Advances in Neural Information Processing Systems, pp. 2672–2680 (2014)
Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., Hochreiter, S.: GANs trained by a two time-scale update rule converge to a local nash equilibrium. In: Advances in Neural Information Processing Systems, pp. 6626–6637 (2017)
Karras, T., Aila, T., Laine, S., Lehtinen, J.: Progressive growing of GANs for improved quality, stability, and variation. In: International Conference on Learning Representations (2018)
Kilcher, Y., Lucchi, A., Hofmann, T.: Semantic interpolation in implicit models. In: International Conference on Learning Representations (2018)
Kingma, D.P., Ba, J.: Adam: A method for stochastic optimization. In: 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, 7–9 May 2015, Conference Track Proceedings (2015)
Kingma, D.P., Welling, M.: Auto-encoding variational Bayes. In: 2nd International Conference on Learning Representations, ICLR 2014, Banff, AB, Canada, 14–16 April 2014, Conference Track Proceedings (2014)
Kingma, D.P., Welling, M., et al.: An introduction to variational autoencoders. Found. Trends® Mach. Learn. 12(4), 307–392 (2019)
Kingma, D.P., Dhariwal, P.: Glow: generative flow with invertible 1 x 1 convolutions. In: Advances in Neural Information Processing Systems, pp. 10215–10224 (2018)
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)
Krizhevsky, A., Hinton, G.: Learning multiple layers of features from tiny images. Technical report (2009)
LeCun, Y., Cortes, C., Burges, C.: Mnist handwritten digit database. ATT Labs [Online]. http://yann.lecun.com/exdb/mnist, February 2010
Papamakarios, G., Nalisnick, E., Rezende, D.J., Mohamed, S., Lakshminarayanan, B.: Normalizing flows for probabilistic modeling and inference. J. Mach. Learn. Res. 22(57), 1–64 (2021)
Paszke, A., et al.: Pytorch: an imperative style, high-performance deep learning library. In: Advances in Neural Information Processing Systems, pp. 8024–8035 (2019)
Radford, A., Metz, L., Chintala, S.: Unsupervised representation learning with deep convolutional generative adversarial networks. In: International Conference on Learning Representations, ICLR 2016 (2016)
Rezende, D., Mohamed, S.: Variational inference with normalizing flows. In: International Conference on Machine Learning, pp. 1530–1538. PMLR (2015)
Rezende, D.J., et al.: Normalizing flows on tori and spheres. In: International Conference on Machine Learning, pp. 8083–8092 (2020)
Rippel, O., Adams, R.P.: High-dimensional probability estimation with deep density models. CoRR abs/1302.5125 (2013)
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)
Shoemake, K.: Animating rotation with quaternion curves. In: Proceedings of the 12th Annual Conference on Computer Graphics and Interactive Techniques, pp. 245–254 (1985)
Tabak, E.G., Turner, C.V.: A family of nonparametric density estimation algorithms. Commun. Pure Appl. Math. 66(2), 145–164 (2013)
Tabak, E.G., Vanden-Eijnden, E.: Density estimation by dual ascent of the log-likelihood. Commun. Pure Appl. Math. 8(1), 217–233 (2010)
White, T.: Sampling generative networks. CoRR abs/1609.04468 (2016)
Xiao, H., Rasul, K., Vollgraf, R.: Fashion-MNIST: a novel image dataset for benchmarking machine learning algorithms. CoRR abs/1708.07747 (2017)
Xu, J., Durrett, G.: Spherical latent spaces for stable variational autoencoders. In: Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pp. 4503–4513 (2018)
Acknowledgements
This research was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil, Finance Code 001 and by FAPESP (grants 2017/24005-2 and 2018/19350-5).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
1 Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Fadel, S.G., Mair, S., da S. Torres, R., Brefeld, U. (2021). Principled Interpolation in Normalizing Flows. In: Oliver, N., Pérez-Cruz, F., Kramer, S., Read, J., Lozano, J.A. (eds) Machine Learning and Knowledge Discovery in Databases. Research Track. ECML PKDD 2021. Lecture Notes in Computer Science(), vol 12976. Springer, Cham. https://doi.org/10.1007/978-3-030-86520-7_8
Download citation
DOI: https://doi.org/10.1007/978-3-030-86520-7_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-86519-1
Online ISBN: 978-3-030-86520-7
eBook Packages: Computer ScienceComputer Science (R0)