Corpus ID: 52889459

Large Scale GAN Training for High Fidelity Natural Image Synthesis

@article{Brock2019LargeSG,
  title={Large Scale GAN Training for High Fidelity Natural Image Synthesis},
  author={Andrew Brock and Jeff Donahue and Karen Simonyan},
  journal={ArXiv},
  year={2019},
  volume={abs/1809.11096}
}
Despite recent progress in generative image modeling, successfully generating high-resolution, diverse samples from complex datasets such as ImageNet remains an elusive goal. [...] Key Method Our modifications lead to models which set the new state of the art in class-conditional image synthesis. When trained on ImageNet at 128x128 resolution, our models (BigGANs) achieve an Inception Score (IS) of 166.5 and Frechet Inception Distance (FID) of 7.4, improving over the previous best IS of 52.52 and FID of 18.6.Expand
Auto-Embedding Generative Adversarial Networks For High Resolution Image Synthesis
TLDR
A novel GAN is developed called auto-embedding generative adversarial network, which simultaneously encodes the global structure features and captures the fine-grained details of real images, and is able to produce high-resolution images of promising quality directly from the input noise. Expand
Exploiting GAN Internal Capacity for High-Quality Reconstruction of Natural Images
TLDR
This work proposes to exploit the representation in intermediate layers of the GAN generator, and shows that this leads to increased capacity and preliminary results in exploiting the learned representation in the attention map of the generator to obtain an unsupervised segmentation of natural images. Expand
High-Fidelity Image Generation With Fewer Labels
TLDR
This work demonstrates how one can benefit from recent work on self- and semi-supervised learning to outperform the state of the art on both unsupervised ImageNet synthesis, as well as in the conditional setting. Expand
Image Processing Using Multi-Code GAN Prior
TLDR
A novel approach is proposed, called mGANprior, to incorporate the well-trained GANs as effective prior to a variety of image processing tasks, by employing multiple latent codes to generate multiple feature maps at some intermediate layer of the generator and composing them with adaptive channel importance to recover the input image. Expand
Adversarial Video Generation on Complex Datasets
TLDR
This work shows that large Generative Adversarial Networks trained on the complex Kinetics-600 dataset are able to produce video samples of substantially higher complexity and fidelity than previous work. Expand
Generating Diverse High-Fidelity Images with VQ-VAE-2
TLDR
It is demonstrated that a multi-scale hierarchical organization of VQ-VAE, augmented with powerful priors over the latent codes, is able to generate samples with quality that rivals that of state of the art Generative Adversarial Networks on multifaceted datasets such as ImageNet, while not suffering from GAN's known shortcomings such as mode collapse and lack of diversity. Expand
ON COMPLEX DATASETS
Generative models of natural images have progressed towards high fidelity samples by the strong leveraging of scale. We attempt to carry this success to the field of video modeling by showing thatExpand
HIGH-FIDELITY FEW-SHOT IMAGE SYNTHESIS
Training Generative Adversarial Networks (GAN) on high-fidelity images usually requires large-scale GPU-clusters and a vast number of training images. In this paper, we study the few-shot imageExpand
Towards Faster and Stabilized GAN Training for High-fidelity Few-shot Image Synthesis
TLDR
This paper proposes a light-weight GAN structure that gains superior quality on 1024 × 1024 resolution, and shows its superior performance compared to the state-of-the-art StyleGAN2, when data and computing budget are limited. Expand
Improved Image Generation via Sparse Modeling
TLDR
It is shown that generators can be viewed as manifestations of the Convolutional Sparse Coding and its Multi-Layered version (ML-CSC) synthesis processes, and explicitly enforcing a sparsifying regularization on appropriately chosen activation layers in the generator, and this leads to improved image synthesis. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 56 REFERENCES
Conditional Image Synthesis with Auxiliary Classifier GANs
TLDR
A variant of GANs employing label conditioning that results in 128 x 128 resolution image samples exhibiting global coherence is constructed and it is demonstrated that high resolution samples provide class information not present in low resolution samples. Expand
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. Expand
Neural Photo Editing with Introspective Adversarial Networks
TLDR
The Neural Photo Editor is presented, an interface that leverages the power of generative neural networks to make large, semantically coherent changes to existing images, and the Introspective Adversarial Network is introduced, a novel hybridization of the VAE and GAN. Expand
Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks
TLDR
A generative parametric model capable of producing high quality samples of natural images using a cascade of convolutional networks within a Laplacian pyramid framework to generate images in a coarse-to-fine fashion. Expand
Megapixel Size Image Creation using Generative Adversarial Networks
TLDR
This work presents an optimized image generation process based on a Deep Convolutional Generative Adversarial Networks (DCGANs) in order to create photorealistic high-resolution images (up to 1024x1024 pixels). Expand
Comparing Generative Adversarial Network Techniques for Image Creation and Modification
TLDR
This paper compares various GAN techniques, both supervised and unsupervised, and adds an encoder to the network, making it possible to encode images to the latent space of the GAN. Expand
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. Expand
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. Expand
On the Quantitative Analysis of Decoder-Based Generative Models
TLDR
This work proposes to use Annealed Importance Sampling for evaluating log-likelihoods for decoder-based models and validate its accuracy using bidirectional Monte Carlo, and analyzes the performance of decoded models, the effectiveness of existing log- likelihood estimators, the degree of overfitting, and the degree to which these models miss important modes of the data distribution. Expand
Improved Training of Wasserstein GANs
TLDR
This work proposes an alternative to clipping weights: penalize the norm of gradient of the critic with respect to its input, which performs better than standard WGAN and enables stable training of a wide variety of GAN architectures with almost no hyperparameter tuning. Expand
...
1
2
3
4
5
...