Graph-Based Contextual Attention Network for Single Image Deraining.

Bin Hu Mingcen Gu Yuehua Li Lili Zhao Shi Cheng

Published in: ICONIP (3) (2022)

Keyphrases

single image
multiple images
shape recovery
d scene
light source
partially occluded
image pairs
shape from shading
super resolution
spherical harmonics
shadow removal
structured light
manhattan world
three dimensional
intrinsic images
scene geometry
high resolution
multiscale