knazeri/edge-connect
EdgeConnect: Structure Guided Image Inpainting using Edge Prediction, ICCV 2019 https://arxiv.org/abs/1901.00212
A two-stage adversarial architecture that decouples inpainting into edge hallucination and image completion, using Canny or external edge detection as structural guidance for missing regions. The framework supports three-stage training (edge model → inpainting model → joint refinement) and integrates with PyTorch on CUDA, accepting irregular masks from multiple datasets (Places2, CelebA, Paris Street-View) with evaluation metrics including PSNR, SSIM, and FID scores.
2,616 stars. No commits in the last 6 months.
Stars
2,616
Forks
533
Language
Python
License
—
Category
Last pushed
Feb 03, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/knazeri/edge-connect"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related models
pathak22/context-encoder
[CVPR 2016] Unsupervised Feature Learning by Image Inpainting using GANs
shepnerd/inpainting_gmcnn
Image Inpainting via Generative Multi-column Convolutional Neural Networks, NeurIPS2018
youyuge34/Anime-InPainting
An application tool of edge-connect, which can do anime inpainting and drawing. 动漫人物图片自动修复，去马赛克，填补，去瑕疵
otenim/GLCIC-PyTorch
A High-Quality PyTorch Implementation of "Globally and Locally Consistent Image Completion".
kkakkkka/HunyuanPortraitLCM
[CVPR-2025] The official code of HunyuanPortrait: Implicit Condition Control for Enhanced...