Title: Super-FAN: Integrated facial landmark localization and super-resolution of
real-world low resolution faces in arbitrary poses with GANs
CVPR 2018 spotlight
Addresses 2 challenging tasks:
- improving the quality of low resolution facial images accurately
- locating the facial landmarks on such poor resolution images
Novelty:
-
Super-FAN(based on GAN): incorporating structural information
integrating a sub-network forintegrating a sub-network for
face alignment through heatmap regression and optimizing a novel
heatmap loss -
benefit:not only on frontal images(正脸) but on the whole spectrum of facial poses
not only on synthetic low resolution images (as in prior work) but also on real-world images.