Corpus ID: 10894094

Improved Training of Wasserstein GANs

@inproceedings{Gulrajani2017ImprovedTO,
  title={Improved Training of Wasserstein GANs},
  author={Ishaan Gulrajani and F. Ahmed and Mart{\'i}n Arjovsky and Vincent Dumoulin and Aaron C. Courville},
  booktitle={NIPS},
  year={2017}
}
  • Ishaan Gulrajani, F. Ahmed, +2 authors Aaron C. Courville
  • Published in NIPS 2017
  • Computer Science, Mathematics
  • Generative Adversarial Networks (GANs) are powerful generative models, but suffer from training instability. [...] Key Method We propose an alternative to clipping weights: penalize the norm of gradient of the critic with respect to its input. Our proposed method performs better than standard WGAN and enables stable training of a wide variety of GAN architectures with almost no hyperparameter tuning, including 101-layer ResNets and language models over discrete data. We also achieve high quality generations on…Expand Abstract
    Progressive Growing of GANs for Improved Quality, Stability, and Variation
    • 2,087
    • Highly Influenced
    • PDF
    Stabilizing Training of Generative Adversarial Networks through Regularization
    • 223
    • Highly Influenced
    • PDF
    GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
    • 1,759
    • PDF
    Spectral Normalization for Generative Adversarial Networks
    • 1,433
    • Highly Influenced
    • PDF
    mixup: Beyond Empirical Risk Minimization
    • 1,013
    • PDF
    Wasserstein Auto-Encoders
    • 332
    • PDF
    A Style-Based Generator Architecture for Generative Adversarial Networks
    • 1,161
    • Highly Influenced
    • PDF
    Self-Attention Generative Adversarial Networks
    • 907
    • PDF
    Large Scale GAN Training for High Fidelity Natural Image Synthesis
    • 1,194
    • Highly Influenced
    • PDF
    Gradient descent GAN optimization is locally stable
    • 189
    • PDF

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 41 REFERENCES
    Generative Adversarial Nets
    • 17,884
    • PDF
    Improved Techniques for Training GANs
    • 3,603
    • PDF
    Least Squares Generative Adversarial Networks
    • 1,578
    • PDF
    BEGAN: Boundary Equilibrium Generative Adversarial Networks
    • 729
    • PDF
    Unrolled Generative Adversarial Networks
    • 503
    • PDF
    f-GAN: Training Generative Neural Samplers using Variational Divergence Minimization
    • 766
    • PDF
    Adam: A Method for Stochastic Optimization
    • 50,046
    • PDF
    Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
    • 6,573
    • Highly Influential
    • PDF
    Boundary-Seeking Generative Adversarial Networks
    • 94
    • PDF
    SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient
    • 1,110
    • PDF