SOTAVerified

Facial Inpainting

Facial inpainting (or face completion) is the task of generating plausible facial structures for missing pixels in a face image.

( Image credit: SymmFCNet )

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–25 of 60 papers

TitleStatusHype
Face Swap via Diffusion ModelCode2
Reference-Guided Large-Scale Face Inpainting with Identity and Texture ControlCode1
ABANet: Attention boundary-aware network for image segmentationCode1
Masked Face Inpainting Through Residual Attention UNetCode1
SFI-Swin: Symmetric Face Inpainting with Swin Transformer by Distinctly Learning Face Components DistributionsCode1
SC-FEGAN: Face Editing Generative Adversarial Network with User's Sketch and ColorCode1
Image Fine-grained InpaintingCode1
Non-Deterministic Face Mask Removal Based On 3D PriorsCode1
Conffusion: Confidence Intervals for Diffusion ModelsCode1
PATMAT: Person Aware Tuning of Mask-Aware Transformer for Face InpaintingCode1
Reference Guided Image Inpainting using Facial AttributesCode1
Do Inpainting Yourself: Generative Facial Inpainting Guided by ExemplarsCode1
E4S: Fine-grained Face Swapping via Editing With Regional GAN InversionCode1
Does Generative Face Completion Help Face Recognition?Code0
FSGAN: Subject Agnostic Face Swapping and ReenactmentCode0
Detecting Overfitting of Deep Generative Networks via Latent RecoveryCode0
LaFIn: Generative Landmark Guided Face InpaintingCode0
Learning Symmetry Consistent Deep CNNs for Face CompletionCode0
Generative Face CompletionCode0
Face Mask Removal with Region-attentive Face InpaintingCode0
Compensation Sampling for Improved Convergence in Diffusion ModelsCode0
FCSR-GAN: Joint Face Completion and Super-resolution via Multi-task LearningCode0
HiMFR: A Hybrid Masked Face Recognition Through Face InpaintingCode0
E2F-Net: Eyes-to-Face Inpainting via StyleGAN Latent SpaceCode0
Domain Embedded Multi-model Generative Adversarial Networks for Image-based Face Inpainting—0
Show:102550
← PrevPage 1 of 3Next →

No leaderboard results yet.