Understanding the Effectiveness of Lipschitz-Continuity in Generative Adversarial Nets
Explore this paper's citation graph
Summary
It is proved in this paper that Lipschitz-continuity condition is a general solution to make the gradient of the optimal discriminative function reliable and leads to a broad family of valid GAN objectives under LipsChitz-Continuity condition, where Wasserstein distance is one special case.
- Type
- preprint
- Published
- 2018-07-02
- Cited by
- 9
- References
- 46
- Access
- Open access
- OpenAlex
- https://openalex.org/W2893744206
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:52903121
Keywords
Lipschitz continuity, Discriminative model, Discriminator, Convergence (economics), Mathematics
References
- Online Learning and Online Convex Optimization
- Optimal Transportation: Continuous and Discrete
- Improved Techniques for Training GANs
- Neural Photo Editing with Introspective Adversarial Networks
- Image-to-Image Translation with Conditional Adversarial Networks
- Unrolled Generative Adversarial Networks
- NIPS 2016 Tutorial: Generative Adversarial Networks
- Towards Principled Methods for Training Generative Adversarial Networks
- Least Squares Generative Adversarial Networks
- Generalization and Equilibrium in Generative Adversarial Nets (GANs)
- On Convergence and Stability of GANs
- The Numerics of GANs
- Do GANs actually learn the distribution? An empirical study
- GANs Trained by a Two Time-Scale Update Rule Converge to a Nash Equilibrium
- Searching for Activation Functions
- Large Scale Optimal Transport and Mapping Estimation
- Solving Approximate Wasserstein GANs to Stationarity
- Which Training Methods for GANs do actually Converge?
- Self-Attention Generative Adversarial Networks
- Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At Every Step
Cited by
- Self-Supervised GAN to Counter Forgetting
- Adversarial Networks and Autoencoders: The Primal-Dual Relationship and Generalization Bounds
- General Probabilistic Surface Optimization and Log Density Estimation
- A Primal-Dual link between GANs and Autoencoders
- Improving GAN Training with Probability Ratio Clipping and Sample Reweighting
- Improved Skin Disease Classification Using Generative Adversarial Network
- Automatic Target Recognition for Low Resolution Foliage Penetrating SAR Images Using CNNs and GANs
- Simultaneous Gradient Descent-Ascent for GANs Minimax Optimization using Sinkhorn Divergence
- Detail Me More: Improving GAN’s photo-realism of complex scenes
Related papers
- Discriminator-Quality Evaluation GAN
- A high-performance, low-cost, leading edge discriminator
- Analysis and computer simulation of a delay discriminator
- Dynamically Masked Discriminator for Generative Adversarial Networks
- Design and Implementation of the Discriminator with Half Approximation Structure in Code Tracking Loop for GPS
- Private GANs, Revisited
- A Gaussian quantum state discriminator
- A Hyperspectral Image Classification Method Based on Multi-Discriminator Generative Adversarial Networks
- A novel pulse shape discriminator
- Mixed batches and symmetric discriminators for GAN training