not-so-BigGAN: Generating High-Fidelity Images on Small Compute with Wavelet-based Super-Resolution

2020-09-09 · Seungwook Han, Akash Srivastava, Cole Hurwitz, Prasanna Sattigeri, David D. Cox

State-of-the-art models for high-resolution image generation, such as BigGAN and VQVAE-2, require an incredible amount of compute resources and/or time (512 TPU-v3 cores) to train, putting them out of reach for the larger research community. On the other hand, GAN-based image super-resolution models, such as ESRGAN, can not only upscale images to high dimensions, but also are efficient to train. In this paper, we present 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. First, we generate images in low-frequency bands by training a sampler in the wavelet domain. Then, we super-resolve these images from the wavelet domain back to the pixel-space with our novel wavelet super-resolution decoder network. Wavelet-based down-sampling method preserves more structural information than pixel-based methods, leading to significantly better generative quality of the low-resolution sampler (e.g., 64x64). Since the sampler and decoder can be trained in parallel and operate on much lower dimensional spaces than end-to-end models, the training cost is substantially reduced. On ImageNet 512x512, our model achieves a Fr\'echet Inception Distance (FID) of 10.59 -- beating the baseline BigGAN model -- at half the compute (256 TPU-v3 cores).

📄 PDF Abstract BibTeX arXiv:2009.04433

Code (0)

등록된 구현이 없습니다.

Tasks

Conditional Image Generation Decoder Image Generation Image Super-Resolution Super-Resolution

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…

Six Ways To Communicate To Someone At Expedia Via Phone And Email's. To communicate or get human at Expedia, the quickest option is typically to call their customer service at ＋１－８８８－８２９－０８８１　ｏｒ　＋１（８０５）　３３０　（４０５６）. You can also use the live chat…

Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…

((Reservation@Faqs))How do I cancel a reservation on Expedia? How do I cancel a reservation on Expedia? +1^888^829^0881° oR +1^888^829^0881 – Need to cancel your Expedia reservation quickly and without hassle? This step-by-step guide…

TTUR The Two Time-scale Update Rule (TTUR) is an update rule for generative adversarial networks trained with stochastic gradient descent. TTUR has an individual learning rate for…

Non-Local Operation A Non-Local Operation is a component for capturing long-range dependencies with deep neural networks. It is a generalization of the classical non-local mean operation in…

Projection Discriminator A Projection Discriminator is a type of discriminator for generative adversarial networks. It is motivated by a probabilistic model in which the distribution of the…

Feedforward Network A Feedforward Network, or a Multilayer Perceptron (MLP), is a neural network with solely densely connected layers. This is the classic neural network architecture of the…

not-so-big-GAN: Generating High-Fidelity Images on Small Compute with Wavelet-based Super-Resolution

2021-01-01 · Seungwook Han, Akash Srivastava, Cole Lincoln Hurwitz, Prasanna Sattigeri 외

Decoder Image Generation Image Super-Resolution Super-Resolution

BigGAN-based Bayesian reconstruction of natural images from human brain activity

2020-03-13 · Kai Qiao, Jian Chen, Linyuan Wang, Chi Zhang 외

In the visual decoding domain, visually reconstructing presented images given the corresponding human brain activity monitored by functional magnetic resonance imaging (fMRI) is difficult, especially when reconstructing …

Conditional Image Generation Generative Adversarial Network Image Generation

Cascaded Diffusion Models for High Fidelity Image Generation

2021-05-30 · Jonathan Ho, Chitwan Saharia, William Chan, David J. Fleet 외

We show that cascaded diffusion models are capable of generating high fidelity images on the class-conditional ImageNet generation benchmark, without any assistance from auxiliary image classifiers to boost sample qualit…

Data Augmentation Image Generation Super-Resolution Vocal Bursts Intensity Prediction

DGL-GAN: Discriminator Guided Learning for GAN Compression

2021-12-13 · Yuesong Tian, Li Shen, Xiang Tian, DaCheng Tao 외

Generative Adversarial Networks (GANs) with high computation costs, e.g., BigGAN and StyleGAN2, have achieved remarkable results in synthesizing high-resolution images from random noise. Reducing the computation cost of …

A cost-effective method for improving and re-purposing large, pre-trained GANs by fine-tuning their class-embeddings

2019-10-10 · Qi Li, Long Mai, Michael A. Alcorn, Anh Nguyen

Large, pre-trained generative models have been increasingly popular and useful to both the research and wider communities. Specifically, BigGANs a class-conditional Generative Adversarial Networks trained on ImageNet---a…

Diversity Model Editing