GAN_stability

GAN evaluation

An open-source project that investigates the stability of Generative Adversarial Networks (GANs) and provides experiments to evaluate their convergence.

Code for paper "Which Training Methods for GANs do actually Converge? (ICML 2018)"

GitHub

918 stars
22 watching
114 forks
Language: Jupyter Notebook
last commit: about 7 years ago

Related projects:

RepositoryDescriptionStars
stormraiser/gan-weight-normImproves the performance of Generative Adversarial Networks by normalizing weights and batch data181
csailvision/gandissectVisualizes and understands internal representations of generative adversarial networks (GANs) to analyze their behavior and improve performance.1,770
tensorflow/ganProvides tools and utilities for training and evaluating Generative Adversarial Networks (GANs) in TensorFlow944
mailmahee/pytorch-generative-adversarial-networksAn implementation of Generative Adversarial Networks in PyTorch for generating realistic data from a given distribution.32
subeeshvasu/awsome-gan-trainingA curated collection of resources and techniques for training Generative Adversarial Networks (GANs) with various architectures, normalization methods, and regularization strategies.27
zcemycl/matlab-ganA collection of MATLAB implementations for Generative Adversarial Networks (GANs) and related deep learning techniques190
mit-han-lab/data-efficient-gansImproves GAN training efficiency by incorporating data augmentation1,286
carpedm20/discogan-pytorchA PyTorch implementation of a Generative Adversarial Network (GAN) for discovering cross-domain relations.1,084
hanzhanggit/stackgan-v2Software framework for training and evaluating generative models of images and text based on adversarial networks847
caogang/wgan-gpAn implementation of an improved training algorithm for generative adversarial networks (GANs), specifically designed to stabilize the training process using a gradient penalty term.1,515
nashory/gans-collection.torchA collection of Torch implementations for training various types of Generative Adversarial Networks (GANs)55
tensorlayer/dcganA deep learning implementation of a generative adversarial network for generating images from text descriptions.340
tolstikhin/adaganAn implementation of an algorithm to train mixtures of Generative Adversarial Networks (GANs) for generative modeling tasks76
justinpinkney/data-efficient-gansAn implementation of Differentiable Augmentation for GAN training to improve data efficiency.11
0b01/simgan-captchaAutomated solution of CAPTCHA images using simulated unsupervised learning and Generative Adversarial Networks (GANs)434