ZPdesu/SEAN
SEAN: Image Synthesis with Semantic Region-Adaptive Normalization (CVPR 2020, Oral)
Introduces semantic region-adaptive normalization (SEAN) for per-region style control in mask-conditioned image synthesis, enabling independent style transfer to different facial regions via reference images. Built on PyTorch with a GAN architecture that encodes region-specific style codes, supporting interactive editing through a PyQt5 UI for real-time mask and style manipulation. Trained on CelebA-HQ/CelebAMask-HQ with multi-GPU support (4 V100s recommended) and includes pretrained models for immediate inference and style interpolation across regions.
658 stars. No commits in the last 6 months.
Stars
658
Forks
92
Language
Python
License
—
Category
Last pushed
Jun 02, 2021
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/ZPdesu/SEAN"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
PaddlePaddle/PaddleGAN
PaddlePaddle GAN library, including lots of interesting applications like First-Order motion...
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...
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