TLWSR: Weakly supervised real-world scene text image super-resolution using text label.

Qin Shi Yu Zhu Chuantao Fang Dawei Yang

Published in: IET Image Process. (2023)

Keyphrases

super resolution
low resolution
high resolution
text mining
viewpoint
higher order
image processing
high quality
video sequences
ground truth
motion estimation
unsupervised learning
image patches
object class