mlsanityData to 7 Oct 2026

On Unsupervised Image-to-image translation and GAN stability

arXiv 2403.09646 ↗First posted 18 Oct 2023cs.CV cs.AI cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

The problem of image-to-image translation is one that is intruiging and challenging at the same time, for the impact potential it can have on a wide variety of other computer vision applications like colorization, inpainting, segmentation and others. Given the high-level of sophistication needed to extract patterns from one domain and successfully applying them to another, especially, in a completely unsupervised (unpaired) manner, this problem has gained much attention as of the last few years. It is one of the first problems where successful applications to deep generative models, and especially Generative Adversarial Networks achieved astounding results that are actually of realworld impact, rather than just a show of theoretical prowess; the such that has been dominating the GAN world. In this work, we study some of the failure cases of a seminal work in the field, CycleGAN [1] and hypothesize that they are GAN-stability related, and propose two general models to try to alleviate these problems. We also reach the same conclusion of the problem being ill-posed that has been also circulating in the literature lately.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 46 of its 59 references that are in the map. First posted on arXiv 18 Oct 2023.

Builds on

46 of 59 references.

References · 46

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2016Densely Connected Convolutional Networkscited by 6,691
  4. 2016Image-to-Image Translation with Conditional Adversarial Networkscited by 4,510
  5. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  6. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,590
  7. 2018Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentationcited by 3,210
  8. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,320
  9. 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,277
  10. 2015Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networkscited by 2,048
  11. 2016Instance Normalization: The Missing Ingredient for Fast Stylizationcited by 1,187
  12. 2017CyCADA: Cycle-Consistent Adversarial Domain Adaptationcited by 1,079
34 more
  1. 2017StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translationcited by 1,036
  2. 2014Striving for Simplicity: The All Convolutional Netcited by 983
  3. 2018Multimodal Unsupervised Image-to-Image Translationcited by 795
  4. 2017Unsupervised Image-to-Image Translation Networkscited by 602
  5. 2017Self-Normalizing Neural Networkscited by 591
  6. 2017Towards Principled Methods for Training Generative Adversarial Networkscited by 460
  7. 2018Understanding disentangling in β-VAEcited by 332
  8. 2017DualGAN: Unsupervised Dual Learning for Image-to-Image Translationcited by 310
  9. 2017Learning Generative Models with Sinkhorn Divergencescited by 310
  10. 2018Diverse Image-to-Image Translation via Disentangled Representationscited by 296
  11. 2017Toward Multimodal Image-to-Image Translationcited by 295
  12. 2016Learning from Simulated and Unsupervised Images through Adversarial Trainingcited by 287
  13. 2017How to Escape Saddle Points Efficientlycited by 284
  14. 2017Learning to Discover Cross-Domain Relations with Generative Adversarial Networkscited by 278
  15. 2017VAE with a VampPriorcited by 247
  16. 2016Coupled Generative Adversarial Networkscited by 216
  17. 2018Improving GANs Using Optimal Transportcited by 122
  18. 2017Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networkscited by 107
  19. 2018Augmented CycleGAN: Learning Many-to-Many Mappings from Unpaired Datacited by 76
  20. 2017Gradient Descent Can Take Exponential Time to Escape Saddle Pointscited by 74
  21. 2017Accelerated Gradient Descent Escapes Saddle Points Faster than Gradient Descentcited by 66
  22. 2017Gradient descent GAN optimization is locally stablecited by 65
  23. 2018Unsupervised Attention-guided Image to Image Translationcited by 58
  24. 2017DiracNets: Training Very Deep Neural Networks Without Skip-Connectionscited by 47
  25. 2018Distribution Matching in Variational Inferencecited by 35
  26. 2018Attention-GAN for Object Transfiguration in Wild Imagescited by 32
  27. 2017Wasserstein Divergence for GANscited by 22
  28. 2018DA-GAN: Instance-level Image Translation by Deep Attention Generative Adversarial Networks (with Supplementary Materials)cited by 21
  29. 2017Memory-Efficient Implementation of DenseNetscited by 17
  30. 2018Network Decoupling: From Regular to Depthwise Separable Convolutionscited by 17
  31. 2018MAGAN: Aligning Biological Manifoldscited by 15
  32. 2018Super-Resolution using Convolutional Neural Networks without Any Checkerboard Artifactscited by 11
  33. 2017A Robust Adaptive Stochastic Gradient Method for Deep Learningcited by 5
  34. 2018Risk Bounds for Unsupervised Cross-Domain Mapping with IPMscited by 2

Built on by

Not cited yet.