Training Transformer Models by Wavelet Losses Improves Quantitative and Visual Performance in Single Image Super-Resolution.

Cansu Korkmaz, A. Murat Tekalp

Browse the full CVPR paper archive.