Mixed batches and symmetric discriminators for GAN training
Explore this paper's citation graph
Summary
This work proposes a generic permutation-invariant discriminator architecture, which is provably a universal approximator of all symmetric functions and reduces mode collapse in GANs on two synthetic datasets, and obtains good results on the CIFAR10 and CelebA datasets.
- Type
- preprint
- Published
- 2018-06-19
- Cited by
- 39
- References
- 26
- Access
- Open access
- OpenAlex
- https://openalex.org/W2800597160
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:49316791
Keywords
Discriminator, Permutation (music), Computer science, Generator (circuit theory), Invariant (physics)
References
- Further results in multiset processing with neural networks
- Stochastic Backpropagation and Approximate Inference in Deep Generative Models
- Auto-Encoding Variational Bayes
- Approximation by superpositions of a sigmoidal function
- Improved Techniques for Training GANs
- PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
- Permutation-equivariant neural networks applied to dynamics prediction
- GANs Trained by a Two Time-Scale Update Rule Converge to a Nash Equilibrium
- Wasserstein Generative Adversarial Networks
- Training generative neural networks via Maximum Mean Discrepancy optimization
- Generative Moment Matching Networks
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- f-GAN: Training Generative Neural Samplers using Variational Divergence Minimization
- Progressive Growing of GANs for Improved Quality, Stability, and Variation
- Attention is All you Need
- Adam: A Method for Stochastic Optimization
- Deep Sets
- Deep Unsupervised Learning using Nonequilibrium Thermodynamics
- Spectral Normalization for Generative Adversarial Networks
- Improved Training of Wasserstein GANs
Cited by
- Leveraging the Exact Likelihood of Deep Latent Variable Models
- A Simple Baseline Algorithm for Graph Classification
- Coverage and Quality Driven Training of Generative Image Models
- GDPP: Learning Diverse Generations Using Determinantal Point Process
- Adversarial Learning With Knowledge of Image Classification for Improving GANs
- Approches pour l'apprentissage incrémental et la génération des images
- Self-supervised learning of predictive segmentation models from video. (Apprentissage autosupervisé de modèles prédictifs de segmentation à partir de vidéos)
- Adaptive Density Estimation for Generative Models
- Creating Artificial Human Genomes Using Generative Models
- Alleviation of Gradient Exploding in GANs: Fake Can Be Real
- Conditional Activation GAN: Improved Auxiliary Classifier GAN
- Deep learning for population size history inference: design, comparison and combination with approximate Bayesian computation
- Early prediction for mode anomaly in generative adversarial network training: An empirical study
- Unsupervised CT Metal Artifact Learning using Attention-guided beta-CycleGAN
- i-Mix: A Strategy for Regularizing Contrastive Representation Learning
- Creating artificial human genomes using generative neural networks
- Interpolation-Based Contrastive Learning for Few-Label Semi-Supervised Learning
- A deep learning guided memetic framework for graph coloring problems
- Semi-supervised Vision Transformers at Scale
- Randomized Quantization for Data Agnostic Representation Learning
Related papers
- A band Mura detection method based on a new generative adversarial network
- GAN Based State-of-Art Image Colorization
- Unrolled Generative Adversarial Networks
- Saliency detection using adversarial learning networks
- Info-flow Enhanced GANs for Recommender
- A Composite Discriminator for Generative Adversarial Network based Video Super-Resolution
- Image Generation Using Different Models Of Generative Adversarial Network