val-iisc/gSRGAN
[ECCV2022] Source Code for "Improving GANs for Long-Tailed Data through Group Spectral Regularization"
This project helps researchers and developers working with image generation models to create more realistic and diverse images, especially for categories with very few examples. It takes in imbalanced image datasets and outputs improved conditional Generative Adversarial Networks (GANs) capable of generating higher quality images across all categories. This tool is for machine learning engineers and researchers focused on computer vision and generative modeling.
No commits in the last 6 months.
Use this if you are training image generation models (GANs) on datasets where some image categories have significantly fewer examples than others, and you need to improve the quality of generated images for those rare categories.
Not ideal if you are working on image classification, object detection, or other discriminative tasks, as this project focuses specifically on improving image generation.
Stars
16
Forks
3
Language
Python
License
MIT
Category
Last pushed
Oct 02, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/val-iisc/gSRGAN"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
yunjey/domain-transfer-network
TensorFlow Implementation of Unsupervised Cross-Domain Image Generation
taesungp/contrastive-unpaired-translation
Contrastive unpaired image-to-image translation, faster and lighter training than cyclegan (ECCV...
PaddlePaddle/PaddleGAN
PaddlePaddle GAN library, including lots of interesting applications like First-Order motion...
tohinz/ConSinGAN
PyTorch implementation of "Improved Techniques for Training Single-Image GANs" (WACV-21)
sagiebenaim/DistanceGAN
Pytorch implementation of "One-Sided Unsupervised Domain Mapping" NIPS 2017