Collaborative Sampling in Generative Adversarial Networks
arXiv:1902.00813
Abstract
The standard practice in Generative Adversarial Networks (GANs) discards the discriminator during sampling. However, this sampling method loses valuable information learned by the discriminator regarding the data distribution. In this work, we propose a collaborative sampling scheme between the generator and the discriminator for improved data generation. Guided by the discriminator, our approach refines the generated samples through gradient-based updates at a particular layer of the generator, shifting the generator distribution closer to the real data distribution. Additionally, we present a practical discriminator shaping method that can smoothen the loss landscape provided by the discriminator for effective sample refinement. Through extensive experiments on synthetic and image datasets, we demonstrate that our proposed method can improve generated samples both quantitatively and qualitatively, offering a new degree of freedom in GAN sampling.
arXiv commentAccepted to AAAI 2020
References in corpus (27)
- Adam: A Method for Stochastic Optimization
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
- Large Scale GAN Training for High Fidelity Natural Image Synthesis
- Spectral Normalization for Generative Adversarial Networks
- Self-Attention Generative Adversarial Networks
- On Calibration of Modern Neural Networks
- Progressive Growing of GANs for Improved Quality, Stability, and Variation
- Towards Deep Learning Models Resistant to Adversarial Attacks
- Improved Techniques for Training GANs
- Theoretically Principled Trade-off between Robustness and Accuracy
- Deep Image Prior
- VEEGAN: Reducing Mode Collapse in GANs using Implicit Variational Learning
- Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models
- On Convergence and Stability of GANs
- Generalization and Equilibrium in Generative Adversarial Nets (GANs)
- Learning to Learn without Gradient Descent by Gradient Descent
- Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At Every Step
- Unrestricted Adversarial Examples
- Stacked Generative Adversarial Networks
- L4: Practical loss-based stepsize adaptation for deep learning
- Discriminator Rejection Sampling
- Defense Against the Dark Arts: An overview of adversarial example security research and future research directions
- Metropolis-Hastings Generative Adversarial Networks
- Stopping GAN Violence: Generative Unadversarial Networks
- Generative Collaborative Networks for Single Image Super-Resolution
- Phase Collaborative Network for Two-Phase Medical Image Segmentation