Student or educator? Sign up with your school email and get 6 months of Pro free.
Cited by 225

The DOI matches its Crossref record; 16 of the 16 references it cites were found in a registry or an index.

On the Effectiveness of Least Squares Generative Adversarial Networks

Open access
Mao, Xudong, Li, Qing, Xie, Haoran, Lau, Raymond Y.K., Wang, Zhen, Smolley, Stephen Paul · 2019 · IEEE Transactions on Pattern Analysis and Machine Intelligence

Abstract

Unsupervised learning with generative adversarial networks (GANs) has proven to be hugely successful. Regular GANs hypothesize the discriminator as a classifier with the sigmoid cross entropy loss function. However, we found that this loss function may lead to the vanishing gradients problem during the learning process. To overcome such a problem, we propose in this paper the Least Squares Generative Adversarial Networks (LSGANs) which adopt the least squares loss for both the discriminator and the generator. We show that minimizing the objective function of LSGAN yields minimizing the Pearson$\chi ^2$divergence. We also show that the derived objective function that yields minimizing the Pearson$\chi ^2$divergence performs better than the classical one of using least squares for classification. There are two benefits of LSGANs over regular GANs. First, LSGANs are able to generate higher quality images than regular GANs. Second, LSGANs perform more stably during the learning process. For evaluating the image quality, we conduct both qualitative and quantitative experiments, and the experimental results show that LSGANs can generate higher quality images than regular GANs. Furthermore, we evaluate the stability of LSGANs in two groups. One is to compare between LSGANs and regular GANs without gradient penalty. We conduct three experiments, including Gaussian mixture distribution, difficult architectures, and a newly proposed method — datasets with small variability, to illustrate the stability of LSGANs. The other one is to compare between LSGANs with gradient penalty (LSGANs-GP) and WGANs with gradient penalty (WGANs-GP). The experimental results show that LSGANs-GP succeed in training for all the difficult architectures used in WGANs-GP, including 101-layer ResNet.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/tpami.2018.2872043

Cites

16 of 16 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Research works printed but not linked: 32

  1. roth, 2017, Stabilizing training of generative adversarial networks through regularization, Proc Advances Neural Inf Process Syst, 2018
  2. mescheder, 2017, The numerics of GANs, Proc Advances Neural Inf Process Syst, 1825
  3. che, 0, Mode regularized generative adversarial networks, CoRR
  4. 10.1109/CVPR.2017.202 · DOI · Paper page
  5. denton, 2015, Deep generative image models using a laplacian pyramid of adversarial networks, Proc Advances Neural Inf Process Syst, 1486
  6. li, 0, Alice: Towards understanding adversarial learning for joint distribution matching, Adv Neural Inf Process Syst (NIPS), 5495
  7. 10.1109/TIT.2010.2068870 · DOI · Paper page
  8. dai, 0, Calibrating energy-based generative adversarial networks, CoRR
  9. chen, 2016, Infogan: Interpretable representation learning by information maximizing generative adversarial nets, Proc Advances Neural Inf Process Syst, 2172
  10. fedus, 0, Many paths to equilibrium: Gans do not need to decrease a divergence at every step, CoRR
  11. metz, 0, Unrolled generative adversarial networks, CoRR
  12. qi, 0, Loss-sensitive generative adversarial networks on lipschitz densities, CoRR
  13. bishop, 2006, Pattern Recognition and Machine Learning
  14. kodali, 0, On convergence and stability of gans, CoRR
  15. donahue, 0, Adversarial feature learning, CoRR
  16. 10.1109/CVPR.2015.7298965 · DOI · Paper page
  17. kingma, 0, Auto-encoding variational bayes, CoRR
  18. dumoulin, 0, Adversarially learned inference, CoRR
  19. salakhutdinov, 2009, Deep Boltzmann machines, Proc Int Conf Artif Intell Statist, 5, 448
  20. goodfellow, 2016, Deep Learning
  21. goodfellow, 2014, Generative adversarial nets, Proc Advances Neural Inf Process Syst, 2672
  22. 10.1109/CVPR.2017.374 · DOI · Paper page
  23. 10.1109/ICCV.2017.304 · DOI · Paper page
  24. netzer, 2011, Reading digits in natural images with unsupervised feature learning, Proc NIPS Workshop on Deep Learning and Unsupervised Feature Learning
  25. hinton, 2009, Replicated softmax: An undirected topic model, Proc Advances Neural Inf Process Syst, 1607
  26. tieleman, 2012, Lecture 6.5|RMSProp: Divide the gradient by a running average of its recent magnitude, COURSERA Neural Netw Mach Learn, 4, 26
  27. taylor, 2010, Convolutional learning of spatio-temporal features, Proc Eur Conf Comput Vis, 140
  28. dieng, 2017, Variational inference via $\chi$? upper bound minimization, Proc Advances Neural Inf Process Syst, 2732
  29. ganin, 2016, Domain-adversarial training of neural networks, J Mach Learn Res, 17, 2096
  30. huszár, 0, How (not) to train your generative model: Scheduled sampling, likelihood, adversary?, CoRR
  31. zhang, 2008, Cat head detection - how to effectively exploit shape and texture features, Proc Eur Conf Comput Vis, 802
  32. 0, TensorFlow: Large-scale machine learning on heterogeneous systems, CoRR

Cited by

See all 225

Metadata

Authors
Mao, X., Li, Q., Xie, H., Lau, R. Y. K., Wang, Z., & Smolley, S. P.
Venue
IEEE Transactions on Pattern Analysis and Machine Intelligence
Published
2019-12-01
Volume
41
Issue
12
Pages
2947-2960
ISSN
  • 0162-8828 (ISSN-L)
  • 2160-9292
  • 1939-3539
OpenAlex
W2962765321
Cited by
225
References
16
FWCI
13.83
Type
journal article
Language
en
Open access
green
Funders
  • Hong Kong Research Grants Council (CityU 11211417)
  • City University of Hong Kong (9610367)

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.