{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,22]],"date-time":"2026-04-22T07:28:29Z","timestamp":1776842909475,"version":"3.51.2"},"reference-count":40,"publisher":"MDPI AG","issue":"8","license":[{"start":{"date-parts":[[2020,8,13]],"date-time":"2020-08-13T00:00:00Z","timestamp":1597276800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Entropy"],"abstract":"<jats:p>One of the major shortcomings of variational autoencoders is the inability to produce generations from the individual modalities of data originating from mixture distributions. This is primarily due to the use of a simple isotropic Gaussian as the prior for the latent code in the ancestral sampling procedure for data generations. In this paper, we propose a novel formulation of variational autoencoders, conditional prior VAE (CP-VAE), with a two-level generative process for the observed data where continuous z and a discrete c variables are introduced in addition to the observed variables x. By learning data-dependent conditional priors, the new variational objective naturally encourages a better match between the posterior and prior conditionals, and the learning of the latent categories encoding the major source of variation of the original data in an unsupervised manner. Through sampling continuous latent code from the data-dependent conditional priors, we are able to generate new samples from the individual mixture components corresponding, to the multimodal structure over the original data. Moreover, we unify and analyse our objective under different independence assumptions for the joint distribution of the continuous and discrete latent variables. We provide an empirical evaluation on one synthetic dataset and three image datasets, FashionMNIST, MNIST, and Omniglot, illustrating the generative performance of our new model comparing to multiple baselines.<\/jats:p>","DOI":"10.3390\/e22080888","type":"journal-article","created":{"date-parts":[[2020,8,13]],"date-time":"2020-08-13T09:23:44Z","timestamp":1597310624000},"page":"888","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":5,"title":["Data-Dependent Conditional Priors for Unsupervised Learning of Multimodal Data"],"prefix":"10.3390","volume":"22","author":[{"given":"Frantzeska","family":"Lavda","sequence":"first","affiliation":[{"name":"Faculty of Science, Computer Science Department, University of Geneva, 1214 Geneva, Switzerland"},{"name":"Geneva School of Business Administration (DMML Group), HES-SO, 1227 Geneva, Switzerland"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Magda","family":"Gregorov\u00e1","sequence":"additional","affiliation":[{"name":"Geneva School of Business Administration (DMML Group), HES-SO, 1227 Geneva, Switzerland"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Alexandros","family":"Kalousis","sequence":"additional","affiliation":[{"name":"Geneva School of Business Administration (DMML Group), HES-SO, 1227 Geneva, Switzerland"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2020,8,13]]},"reference":[{"key":"ref_1","unstructured":"Kingma, D.P., and Welling, M. (2014, January 14\u201316). Auto-encoding variational bayes. Proceedings of the 2nd International Conference on Learning Representations (ICLR), Banff, AB, Canada."},{"key":"ref_2","unstructured":"Rezende, D.J., Mohamed, S., and Wierstra, D. (2014, January 21\u201326). Stochastic backpropagation and approximate inference in deep generative models. Proceedings of the International Conference on Machine Learning, Beijing, China."},{"key":"ref_3","unstructured":"Rezende, D.J., and Mohamed, S. (2015, January 6\u201311). Variational inference with normalizing flows. Proceedings of the 32 International Conference on Machine Learning, Lille, France."},{"key":"ref_4","unstructured":"Burda, Y., Grosse, R.B., and Salakhutdinov, R. (2016, January 2\u20134). Importance weighted autoencoders. Proceedings of the 4th International Conference on Learning Representations (ICLR), San Juan, Puerto Rico."},{"key":"ref_5","unstructured":"Kingma, D.P., Salimans, T., Jozefowicz, R., Chen, X., Sutskever, I., and Welling, M. (2016, January 5\u201310). Improved variational inference with inverse autoregressive flow. Proceedings of the Advances in Neural Information Processing Systems 2016, Barcelona, Spain."},{"key":"ref_6","unstructured":"Chen, X., Kingma, D.P., Salimans, T., Duan, Y., Dhariwal, P., Schulman, J., Sutskever, I., and Abbeel, P. (2017, January 24\u201326). Variational lossy autoencoders. Proceedings of the International Conference on Learning Representations, Toulon, France."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Jiang, Z., Zheng, Y., Tan, H., Tang, B., and Zhou, H. (2017, January 19\u201325). Variational deep embedding: An unsupervised and generative approach to clustering. Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence, Melbourne, Australia.","DOI":"10.24963\/ijcai.2017\/273"},{"key":"ref_8","unstructured":"Nalisnick, E., and Smyth, P. (2017, January 24\u201326). Stick-breaking variational autoencoders. Proceedings of the International Conference on Learning Representations, Toulon, France."},{"key":"ref_9","unstructured":"Zhao, S., Song, J., and Ermon, S. (2017, January 4\u20139). InfoVAE: Information maximizing variational autoencoders. Proceedings of the Thirty-Third AAAI Conference on Artificial Intelligence (AAAI) 2017, San Francisco, CA, USA."},{"key":"ref_10","unstructured":"Alemi, A.A., Poole, B., Fischer, I., Dillon, J.V., Saurous, R.A., and Murphy, K. (2018, January 10\u201315). Fixing a broken ELBO. Proceedings of the 35th International Conference on Machine Learning (ICML), Stockholm, Sweden."},{"key":"ref_11","unstructured":"Davidson, T.R., Falorsi, L., De Cao, N., Kipf, T., and Tomczak, J.M. (2018). Hyperspherical variational auto-encoders. arXiv."},{"key":"ref_12","unstructured":"Dai, B., and Wipf, D. (2019, January 6\u20139). Diagnosing and enhancing VAE models. Proceedings of the International Conference on Learning Representations, New Orleans, LA, USA."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Ramapuram, J., Gregorova, M., and Kalousis, A. (2019). Lifelong generative modeling. arXiv.","DOI":"10.1016\/j.neucom.2020.02.115"},{"key":"ref_14","unstructured":"Lavda, F., Ramapuram, J., Gregorova, M., and Kalousis, A. (2018). Continual classification learning using generative models. Continual learning workshop, Advances in Neural Information Processing Systems2018, Montreal, QC, Canada, 3\u20138 December 2018. arXiv."},{"key":"ref_15","unstructured":"Locatello, F., Bauer, S., Lucic, M., R\u00e4tsch, G., Gelly, S., Sch\u00f6lkopf, B., and Bachem, O. (2019, January 10\u201315). Challenging common assumptions in the unsupervised learning of disentangled representations. Proceedings of the 36th International Conference on Machine Learning, Long Beach, CA, USA."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Bowman, S.R., Vilnis, L., Vinyals, O., Dai, A., Jozefowicz, R., and Bengio, S. (2016). Generating Sentences from a Continuous Space. Proceedings of the 20th SIGNLL Conference on Computational Natural Language Learning, Association for Computational Linguistics.","DOI":"10.18653\/v1\/K16-1002"},{"key":"ref_17","unstructured":"Tomczak, J.M., and Welling, M. (2018, January 9\u201311). VAE with a VampPrior. Proceedings of the 21st International Conference on Artificial Intelligence and Statistics (AISTATS), Playa Blanca, Lanzarote, Spain."},{"key":"ref_18","unstructured":"Higgins, I., Matthey, L., Pal, A., Burgess, C., Glorot, X., Botvinick, M., Mohamed, S., and Lerchner, A. (2017, January 24\u201326). Beta-VAE: Learning basic visual concepts with a constrained variational framework. Proceedings of the International Conference on Learning Representations (ICLR), Toulon, France."},{"key":"ref_19","unstructured":"Hoffman, M.D., and Johnson, M.J. (2018, January 2). ELBO surgery: Yet another way to carve up the variational evidence lower bound. Proceedings of the NIPS Symposium on Advances in Approximate Bayesian Inference, Montreal, QC, Canada."},{"key":"ref_20","unstructured":"Kim, H., and Mnih, A. (2018). Disentangling by factorising. arXiv."},{"key":"ref_21","unstructured":"Dupont, E. (2018, January 3\u20138). Learning disentangled joint continuous and discrete representations. Proceedings of the Advances in Neural Information Processing Systems 2018, Montreal, QC, Canada."},{"key":"ref_22","unstructured":"Gulrajani, I., Kumar, K., Ahmed, F., Taiga, A.A., Visin, F., Vazquez, D., and Courville, A. (2017, January 24\u201326). PixelVAE: A latent variable model for natural images. Proceedings of the International Conference on Learning Representations (ICLR) 2017, Toulon, France."},{"key":"ref_23","unstructured":"Cortes, C., Lawrence, N.D., Lee, D.D., Sugiyama, M., and Garnett, R. (2015). A recurrent latent variable model for sequential data. Advances in Neural Information Processing Systems 28, Curran Associates, Inc."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Xu, J., and Durrett, G. (November, January 31). Spherical latent spaces for stable variational autoencoders. Proceedings of the Conference on Empirical Methods in Natural Language Processing 2018, Brussels, Belgium.","DOI":"10.18653\/v1\/D18-1480"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Takahashi, H., Iwata, T., Yamanaka, Y., Yamada, M., and Yagi, S. (February, January 27). Variational autoencoder with implicit optimal priors. Proceedings of the AAAI Conference on Artificial Intelligence 2019, Honolulu, HI, USA.","DOI":"10.1609\/aaai.v33i01.33015066"},{"key":"ref_26","unstructured":"Dilokthanakul, N., Mediano, P.A.M., Garnelo, M., Lee, M.C.H., Salimbeni, H., Arulkumaran, K., and Shanahan, M. (2017). Deep unsupervised clustering with gaussian mixture variational autoencoders. arXiv."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Goyal, P., Hu, Z., Liang, X., Wang, C., and Xing, E. (2017, January 22\u201329). Nonparametric variational auto-encoders for hierarchical representation learning. Proceedings of the IEEE International Conference on Computer Vision 2017, Venice, Italy.","DOI":"10.1109\/ICCV.2017.545"},{"key":"ref_28","unstructured":"Li, X., Chen, Z., Poon, L.K.M., and Zhang, N.L. (2019, January 6\u20139). Learning Latent Superstructures in Variational Autoencoders for Deep Multidimensional Clustering. Proceedings of the International Conference on Learning Representations 2019, New Orleans, LA, USA."},{"key":"ref_29","unstructured":"Murphy, K.P. (2012). Machine Learning: A Probabilistic Perspective, MIT Press."},{"key":"ref_30","unstructured":"Kingma, D.P., and Ba, J. (2015, January 7\u20139). Adam: A method for stochastic optimization. Proceedings of the 3rd International Conference for Learning Representations 2015, San Diego, CA, USA."},{"key":"ref_31","unstructured":"Mohamed, S., Rosca, M., Figurnov, M., and Mnih, A. (2019). Monte carlo gradient estimation in machine learning. arXiv."},{"key":"ref_32","unstructured":"Jang, E., Gu, S., and Poole, B. (2017, January 24\u201326). Categorical reparameterization with gumbel-softmax. Proceedings of the International Conference on Learning Representations (ICLR) 2017, Toulon, France."},{"key":"ref_33","unstructured":"Makhzani, A., Shlens, J., Jaitly, N., Goodfellow, I., and Frey, B. (2016). Adversarial autoencoders. arXiv."},{"key":"ref_34","unstructured":"Ghahramani, Z., Welling, M., Cortes, C., Lawrence, N.D., and Weinberger, K.Q. (2014). Semi-supervised learning with deep generative models. Advances in Neural Information Processing Systems 27, Curran Associates, Inc."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","author":"Lecun","year":"1998","journal-title":"Proc. IEEE"},{"key":"ref_36","unstructured":"Xiao, H., Rasul, K., and Vollgraf, R. (2017). Fashion-mnist: A novel image dataset for benchmarking machine learning algorithms. arXiv."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"1332","DOI":"10.1126\/science.aab3050","article-title":"Human-level concept learning through probabilistic program induction","volume":"350","author":"Lake","year":"2015","journal-title":"Science"},{"key":"ref_38","unstructured":"Glorot, X., and Bengio, Y. (2010, January 13\u201315). Understanding the difficulty of training deep feedforward neural networks. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, Sardinia, Italy."},{"key":"ref_39","unstructured":"Dauphin, Y.N., Fan, A., Auli, M., and Grangier, D. (2017, January 6\u201311). Language modeling with gated convolutional networks. Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia."},{"key":"ref_40","doi-asserted-by":"crossref","unstructured":"Salakhutdinov, R., and Murray, I. (2008, January 5\u20139). On the quantitative analysis of deep belief networks. Proceedings of the 25th International Conference on Machine Learning, Helsinki, Finland.","DOI":"10.1145\/1390156.1390266"}],"container-title":["Entropy"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1099-4300\/22\/8\/888\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T10:00:01Z","timestamp":1760176801000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1099-4300\/22\/8\/888"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,8,13]]},"references-count":40,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2020,8]]}},"alternative-id":["e22080888"],"URL":"https:\/\/doi.org\/10.3390\/e22080888","relation":{},"ISSN":["1099-4300"],"issn-type":[{"value":"1099-4300","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,8,13]]}}}