mlsanityData to 7 Oct 2026

Gromov-Wasserstein Autoencoders

arXiv 2209.07007 ↗First posted 15 Sep 2022cs.LG cs.CVcited by 7

Not on the map. Why?

Abstract (arXiv)

Variational Autoencoder (VAE)-based generative models offer flexible representation learning by incorporating meta-priors, general premises considered beneficial for downstream tasks. However, the incorporated meta-priors often involve ad-hoc model deviations from the original likelihood architecture, causing undesirable changes in their training. In this paper, we propose a novel representation learning method, Gromov-Wasserstein Autoencoders (GWAE), which directly matches the latent and data distributions using the variational autoencoding scheme. Instead of likelihood-based objectives, GWAE models minimize the Gromov-Wasserstein (GW) metric between the trainable prior and given data distributions. The GW metric measures the distance structure-oriented discrepancy between distributions even with different dimensionalities, which provides a direct measure between the latent and data spaces. By restricting the prior family, we can introduce meta-priors into the latent space without changing their objective. The empirical comparisons with VAE-based models show that GWAE models work in two prominent meta-priors, disentanglement and clustering, with their GW objective unchanged.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 50 of its 89 references that are in the map. First posted on arXiv 15 Sep 2022.

Builds on

50 of 89 references.

References · 50

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2013Auto-Encoding Variational Bayescited by 3,980
  4. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  5. 2016Gaussian Error Linear Units (GELUs)cited by 3,431
  6. 2017Improved Training of Wasserstein GANscited by 2,223
  7. 2015Variational Inference with Normalizing Flowscited by 2,161
  8. 2018Spectral Normalization for Generative Adversarial Networkscited by 1,618
  9. 2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 929
  10. 2016Deep Variational Information Bottleneckcited by 915
  11. 2020NVAE: A Deep Hierarchical Variational Autoencodercited by 635
  12. 2018Disentangling by Factorisingcited by 614
38 more
  1. 2019Self-labelling via simultaneous clustering and representation learningcited by 525
  2. 2017Towards Principled Methods for Training Generative Adversarial Networkscited by 460
  3. 2015Autoencoding beyond pixels using a learned similarity metriccited by 407
  4. 2017Wasserstein Auto-Encoderscited by 371
  5. 2016Adversarial Feature Learningcited by 350
  6. 2016Ladder Variational Autoencoderscited by 315
  7. 2017Toward Controlled Generation of Textcited by 271
  8. 2017VAE with a VampPriorcited by 247
  9. 2017Variational Inference of Disentangled Latent Concepts from Unlabeled Observationscited by 240
  10. 2016Adversarially Learned Inferencecited by 216
  11. 2019Flexibly Fair Representation Learning by Disentanglementcited by 196
  12. 2019Diagnosing and Enhancing VAE Modelscited by 192
  13. 2017DARLA: Improving Zero-Shot Transfer in Reinforcement Learningcited by 147
  14. 2019On the Fairness of Disentangled Representationscited by 144
  15. 2016Information Dropout: Learning Optimal Representations Through Noisy Computationcited by 131
  16. 2018Recent Advances in Autoencoder-Based Representation Learningcited by 130
  17. 2017Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networkscited by 107
  18. 2020Measuring Disentanglement: A Review of Metricscited by 69
  19. 2016Amortised MAP Inference for Image Super-resolutioncited by 68
  20. 2020A Contrastive Learning Approach for Training Variational Autoencoder Priorscited by 65
  21. 2019Theory and Evaluation Metrics for Learning Disentangled Representationscited by 57
  22. 2020Guided Variational Autoencoder for Disentanglement Learningcited by 57
  23. 2020Simple and Effective VAE Training with Calibrated Decoderscited by 57
  24. 2020The Unbalanced Gromov Wasserstein Distance: Conic Formulation and Relaxationcited by 55
  25. 2019The Usual Suspects? Reassessing Blame for VAE Posterior Collapsecited by 52
  26. 2020A Sober Look at the Unsupervised Learning of Disentangled Representations and their Evaluationcited by 52
  27. 2017Variational Inference using Implicit Distributionscited by 39
  28. 2017Learning Hierarchical Features from Generative Modelscited by 27
  29. 2017Learning Latent Representations for Speech Generation and Transformationcited by 27
  30. 2020Smoothness and Stability in GANscited by 27
  31. 2019Explicit Disentanglement of Appearance and Perspective in Generative Modelscited by 20
  32. 2020Learning Autoencoders with Relational Regularizationcited by 20
  33. 2018Wasserstein Variational Inferencecited by 17
  34. 2018Disentangling the independently controllable factors of variation by interacting with the worldcited by 16
  35. 2020Improving Relational Regularized Autoencoders with Spherical Sliced Fused Gromov Wassersteincited by 13
  36. 2019Improving Variational Autoencoder with Deep Feature Consistent and Generative Adversarial Trainingcited by 12
  37. 2020Learning disentangled representations with the Wasserstein Autoencodercited by 3
  38. 2018Implicit Autoencoderscited by 2

Built on by

Cited by year

Citing papers per year: 2023: 3; 2024: 1; 2025: 3.

7 papers.

  1. 2023Scalable Optimal Transport Methods in Machine Learning: A Contemporary Surveycited by 27
  2. 2023Monotone Generative Modeling via a Gromov-Monge Embeddingcited by 3
  3. 2024Preventing Model Collapse in Gaussian Process Latent Variable Modelscited by 3
  4. 2023Simplex Autoencoderscited by 0
  5. 2025Sparsification and Reconstruction from the Perspective of Representation Geometrycited by 0
  6. 2025Relation-Aware Slicing in Cross-Domain Alignmentcited by 0
  7. 2025TRANSPORTER: Transferring Visual Semantics from VLM Manifoldscited by 0