This repository implements the paper "Reference based Sketch Image Colorization using Augmented-Self Reference and Dense Semantic Correspondence" which is published in CVPR2020.
Can you provide the code for visualization of the attention maps as shown in Figure-7 of the paper or recommend some way for the visulization of the maps as the input to SCFT are feature maps not images. Thanks.
Can you provide the code for visualization of the attention maps as shown in Figure-7 of the paper or recommend some way for the visulization of the maps as the input to SCFT are feature maps not images. Thanks.