Abstract
Previous line of research on learning disentangled representations in an unsupervised setting focused on enforcing an uncorrelated posterior. These approaches have been shown both empirically and theoretically to be insufficient for guaranteeing disentangled representations. Recent works postulate that an implicit PCA-like behavior might explain why these models still tend to disentangle, exploiting the structure of variance in the datasets. Here we aim to further verify those hypotheses by conducting multiple analyses on existing benchmark datasets and models, focusing on the relation between the structure of variance induced by the ground-truth factors and properties of the learned representations. We quantify the effects of global and local directions of variance in the data on disentanglement performance using proposed measures and seem to find empirical evidence of a negative effect of local variance directions on disentanglement. We also invalidate the robustness of models with a global ordering of latent dimensions against the local vs. global discrepancies in the data.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Baldi, P., Hornik, K.: Neural networks and principal component analysis: learning from examples without local minima. Neural Netw. 2(1), 53–58 (1989)
Bengio, Y., Courville, A., Vincent, P.: Representation learning: a review and new perspectives. IEEE Trans. Pattern Anal. Mach. Intell. 35(8), 1798–1828 (2013)
Bouchacourt, D., Tomioka, R., Nowozin, S.: Multi-level variational autoencoder: learning disentangled representations from grouped observations. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 32 (2018)
Bourlard, H., Kamp, Y.: Auto-association by multilayer perceptrons and singular value decomposition. Biol. Cybern. 59(4), 291–294 (1988)
Burgess, C.P., et al.: Understanding disentangling in beta-VAE. arXiv preprint arXiv:1804.03599 (2018)
Chen, T.Q., Li, X., Grosse, R.B., Duvenaud, D.K.: Isolating sources of disentanglement in variational autoencoders. In: Advances in Neural Information Processing Systems, pp. 2610–2620 (2018)
Dai, B., Wang, Y., Aston, J., Hua, G., Wipf, D.: Connections with robust PCA and the role of emergent sparsity in variational autoencoder models. J. Mach. Learn. Res. 19(1), 1573–1614 (2018)
Duan, S., et al.: Unsupervised model selection for variational disentangled representation learning. In: International Conference on Learning Representations (2019)
Eastwood, C., Williams, C.K.: A framework for the quantitative evaluation of disentangled representations (2018)
Eriksson, J., Koivunen, V.: Identifiability and separability of linear ICA models revisited. In: Proceedings of ICA, vol. 2003, pp. 23–27 (2003)
Gondal, M.W., et al.: On the transfer of inductive bias from simulation to the real world: a new disentanglement dataset. Adv. Neural Inf. Process. Syst. 32, 15740–15751 (2019)
Goodfellow, I., Bengio, Y., Courville, A.: Deep Learning. MIT Press, Cambridge (2016)
Higgins, I., et al.: Early visual concept learning with unsupervised deep learning. arXiv preprint arXiv:1606.05579 (2016)
Hosoya, H.: Group-based learning of disentangled representations with generalizability for novel contents. In: IJCAI, pp. 2506–2513 (2019)
Hyvärinen, A.: Survey on independent component analysis (1999)
Khemakhem, I., Kingma, D., Monti, R., Hyvarinen, A.: Variational autoencoders and nonlinear ICA: a unifying framework. In: International Conference on Artificial Intelligence and Statistics, pp. 2207–2217. PMLR (2020)
Kim, H., Mnih, A.: Disentangling by factorising. In: International Conference on Machine Learning, pp. 2649–2658. PMLR (2018)
Kingma, D.P., Welling, M.: Auto-encoding variational Bayes. arXiv preprint arXiv:1312.6114 (2013)
Kumar, A., Sattigeri, P., Balakrishnan, A.: Variational inference of disentangled latent concepts from unlabeled observations. In: International Conference on Learning Representations (2018)
Kunin, D., Bloom, J., Goeva, A., Seed, C.: Loss landscapes of regularized linear autoencoders. In: Chaudhuri, K., Salakhutdinov, R. (eds.) Proceedings of the 36th International Conference on Machine Learning. Proceedings of Machine Learning Research, vol. 97, pp. 3560–3569. PMLR, 09–15 June 2019
LeCun, Y., Huang, F.J., Bottou, L.: Learning methods for generic object recognition with invariance to pose and lighting. In: Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2004. CVPR 2004. vol. 2, pp. II–104. IEEE (2004)
Locatello, F., et al.: Challenging common assumptions in the unsupervised learning of disentangled representations. In: International Conference on Machine Learning, pp. 4114–4124. PMLR (2019)
Locatello, F., et al.: A commentary on the unsupervised learning of disentangled representations. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 34, pp. 13681–13684 (2020)
Locatello, F., et al.: Weakly-supervised disentanglement without compromises. In: International Conference on Machine Learning, pp. 6348–6359. PMLR (2020)
Lucas, J., Tucker, G., Grosse, R.B., Norouzi, M.: Don’t blame the ELBO! A linear VAE perspective on posterior collapse. In: Wallach, H., Larochelle, H., Beygelzimer, A., d’Alché-Buc, F., Fox, E., Garnett, R. (eds.) Advances in Neural Information Processing Systems, vol. 32. Curran Associates, Inc. (2019)
Pandey, A., Fanuel, M., Schreurs, J., Suykens, J.A.: Disentangled representation learning and generation with manifold optimization. arXiv preprint arXiv:2006.07046 (2020)
Pearson, K.: LIII. On lines and planes of closest fit to systems of points in space. London Edinburgh Dublin Philos. Mag. J. Sci. 2(11), 559–572 (1901)
Peters, J., Janzing, D., Schölkopf, B.: Elements of Causal Inference: Foundations and Learning Algorithms. The MIT Press, Cambridge (2017)
Plaut, E.: From principal subspaces to principal components with linear autoencoders. arXiv preprint arXiv:1804.10253 (2018)
Reed, S.E., Zhang, Y., Zhang, Y., Lee, H.: Deep visual analogy-making. In: Advances in Neural Information Processing Systems, pp. 1252–1260 (2015)
Ren, X., Yang, T., Wang, Y., Zeng, W.: Rethinking content and style: exploring bias for unsupervised disentanglement. arXiv preprint arXiv:2102.10544 (2021)
Ridgeway, K., Mozer, M.C.: Learning deep disentangled embeddings with the f-statistic loss. In: Advances in Neural Information Processing Systems, pp. 185–194 (2018)
Rolinek, M., Zietlow, D., Martius, G.: Variational autoencoders pursue PCA directions (by accident). In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 12406–12415 (2019)
Schölkopf, B., Smola, A., Müller, K.-R.: Kernel principal component analysis. In: Gerstner, W., Germond, A., Hasler, M., Nicoud, J.-D. (eds.) ICANN 1997. LNCS, vol. 1327, pp. 583–588. Springer, Heidelberg (1997). https://doi.org/10.1007/BFb0020217
Scholz, M., Vigário, R.: Nonlinear PCA: a new hierarchical approach. In: ESANN, pp. 439–444 (2002)
Shu, R., Chen, Y., Kumar, A., Ermon, S., Poole, B.: Weakly supervised disentanglement with guarantees. In: International Conference on Learning Representations (2019)
Stühmer, J., Turner, R., Nowozin, S.: Independent subspace analysis for unsupervised learning of disentangled representations. In: International Conference on Artificial Intelligence and Statistics, pp. 1200–1210. PMLR (2020)
Suykens, J.A.: Deep restricted kernel machines using conjugate feature duality. Neural Comput. 29(8), 2123–2163 (2017)
Tipping, M.E., Bishop, C.M.: Probabilistic principal component analysis. J. R. Stat. Soc. Ser. B (Stat. Methodol.) 61(3), 611–622 (1999)
Tonin, F., Patrinos, P., Suykens, J.A.: Unsupervised learning of disentangled representations in deep restricted kernel machines with orthogonality constraints. arXiv preprint arXiv:2011.12659 (2020)
Zietlow, D., Rolinek, M., Martius, G.: Demystifying inductive biases for \(beta\)-VAE based architectures. arXiv preprint arXiv:2102.06822 (2021)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Rakowski, A., Lippert, C. (2021). Disentanglement and Local Directions of Variance. 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 12977. Springer, Cham. https://doi.org/10.1007/978-3-030-86523-8_2
Download citation
DOI: https://doi.org/10.1007/978-3-030-86523-8_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-86522-1
Online ISBN: 978-3-030-86523-8
eBook Packages: Computer ScienceComputer Science (R0)