• Corpus ID: 214802456

Feature Quantization Improves GAN Training

@inproceedings{Zhao2020FeatureQI,
  title={Feature Quantization Improves GAN Training},
  author={Yang Zhao and Chengkun Li and Ping Yu and Jianfeng Gao and Changyou Chen},
  booktitle={ICML},
  year={2020}
}
The instability in GAN training has been a long-standing problem despite remarkable research efforts. We identify that instability issues stem from difficulties of performing feature matching with mini-batch statistics, due to a fragile balance between the fixed target distribution and the progressively generated distribution. In this work, we propose Feature Quantization (FQ) for the discriminator, to embed both true and fake data samples into a shared discrete space. The quantized values of… 
Instance Selection for GANs
TLDR
This work proposes a novel approach to improve sample quality: altering the training dataset via instance selection before model training has taken place, which redirects model capacity towards high-density regions, which ultimately improves sample fidelity.
Training GANs with Stronger Augmentations via Contrastive Discriminator
TLDR
This paper proposes a novel way to address the discriminator overfitting issue in GANs by incorporating a recent contrastive representation learning scheme into the GAN discriminator, coined ContraD, and shows that GAns with ContraD consistently improve FID and IS compared to other recent techniques incorporating data augmentations.
not-so-BigGAN: Generating High-Fidelity Images on Small Compute with Wavelet-based Super-Resolution
TLDR
Not-so-big-GAN (nsb-GAN), a simple yet cost-effective two-step training framework for deep generative models (DGMs) of high-dimensional natural images, which achieves a Frechet Inception Distance of 10.59 -- beating the baseline BigGAN model -- at half the compute.
Scalable Balanced Training of Conditional Generative Adversarial Neural Networks on Image Data
We propose a distributed approach to train deep convolutional generative adversarial neural network (DC-CGANs) models. Our method reduces the imbalance between generator and discriminator by
Latent Video Transformer
TLDR
This work addresses the problem of high computational requirements of generative models for videos via modeling the dynamics in a latent space after the transformation of frames into the latent space, and predicts latent representation for the next frames in an autoregressive manner.
Structure-Aware Human-Action Generation
TLDR
This work proposes a variant of GCNs to leverage the powerful self-attention mechanism to adaptively sparsify a complete action graph in the temporal space, and demonstrates the superiority of this method on two standard human action datasets compared with existing methods.
Generative Models as a Robust Alternative for Image Classification: Progress and Challenges
TLDR
Generative models as a Robust Alternative for Image Classification: Progress and Challenges shows progress and challenges in the development of generative models.
Manifold Learning Benefits GANs
TLDR
This paper improves Generative Adversarial Networks by incorporating a manifold learning step into the discriminator, and concludes that locality-constrained non-linear manifolds have the upper hand over linear manifolds due to their non-uniform density and smoothness.
A Unified Generative Adversarial Network Training via Self-Labeling and Self-Attention
TLDR
This work introduces a form of artificial labeling that can incorporate manually defined labels, when available, and induce an alignment between them, and finds that the proposed scheme can even outperform classconditional GANs.
ICLR 2021 Updating Search Space Class-Agnostic Generator Controller Controller Sampling Training & Evaluation Updating Search Space Gain Intuition
  • Computer Science
  • 2020
TLDR
This paper presents a novel idea that adopts NAS to find a distinct architecture for each class of conditional Generative Adversarial Networks, and discovers several insights that are helpful in designing cGAN models.
...
1
2
3
...

References

SHOWING 1-10 OF 49 REFERENCES
Twin Auxilary Classifiers GAN
TLDR
This paper identifies the source of low diversity issue theoretically and proposes Twin Auxiliary Classifiers Generative Adversarial Net (TAC-GAN), a practical solution to the problem that adds a new player that interacts with other players (the generator and the discriminator) in GAN.
On Leveraging Pretrained GANs for Limited-Data Generation
TLDR
It is revealed that low-level filters of both the generator and discriminator of pretrained GANs can be transferred to facilitate generation in a perceptually-distinct target domain with limited training data.
Progressive Growing of GANs for Improved Quality, Stability, and Variation
TLDR
A new training methodology for generative adversarial networks is described, starting from a low resolution, and adding new layers that model increasingly fine details as training progresses, allowing for images of unprecedented quality.
Which Training Methods for GANs do actually Converge?
TLDR
This paper describes a simple yet prototypical counterexample showing that in the more realistic case of distributions that are not absolutely continuous, unregularized GAN training is not always convergent, and extends convergence results to more general GANs and proves local convergence for simplified gradient penalties even if the generator and data distribution lie on lower dimensional manifolds.
Large Scale GAN Training for High Fidelity Natural Image Synthesis
TLDR
It is found that applying orthogonal regularization to the generator renders it amenable to a simple "truncation trick," allowing fine control over the trade-off between sample fidelity and variety by reducing the variance of the Generator's input.
Improved Techniques for Training GANs
TLDR
This work focuses on two applications of GANs: semi-supervised learning, and the generation of images that humans find visually realistic, and presents ImageNet samples with unprecedented resolution and shows that the methods enable the model to learn recognizable features of ImageNet classes.
Self-Attention Generative Adversarial Networks
TLDR
The proposed SAGAN achieves the state-of-the-art results, boosting the best published Inception score from 36.8 to 52.52 and reducing Frechet Inception distance from 27.62 to 18.65 on the challenging ImageNet dataset.
MMD GAN: Towards Deeper Understanding of Moment Matching Network
TLDR
In the evaluation on multiple benchmark datasets, including MNIST, CIFAR- 10, CelebA and LSUN, the performance of MMD-GAN significantly outperforms GMMN, and is competitive with other representative GAN works.
Improving MMD-GAN Training with Repulsive Loss Function
TLDR
This study revisits MMD-GAN that uses the maximum mean discrepancy (MMD) as the loss function for GAN and proposes a repulsive loss function to actively learn the difference among the real data by simply rearranging the terms in MMD.
GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
TLDR
This work proposes a two time-scale update rule (TTUR) for training GANs with stochastic gradient descent on arbitrary GAN loss functions and introduces the "Frechet Inception Distance" (FID) which captures the similarity of generated images to real ones better than the Inception Score.
...
1
2
3
4
5
...