Blind Face Restoration via Multi-head Cross-attention and Generative Priors.

Zhiyong Huo Shanlin Hu

Published in: CISP-BMEI (2023)

Keyphrases

image restoration
head orientation
facial expressions
generative model
focus of attention
head tracking
facial gestures
face images
data driven
unsupervised learning
head motion
human faces
blind image restoration
upper body
facial images
human head
prior model
image processing
maximum likelihood
video sequences