Giters

The-Learning-And-Vision-Atelier-LAVA / ArbSR

[ICCV 2021] Learning A Single Network for Scale-Arbitrary Super-Resolution

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

about figure 1 of your paper

XiaoyuShi97 opened this issue · comments

commented

Hi, I am confused about figure 1 in your paper. It seems that (c) are intermediate feature maps? But what does the values (i.e. range [0, 5000]) mean. And how do you get figures of par (b)? Thx!

commented

Hi @btwbtm, thanks for your interest in our work. The value in Fig. 1 is calculated using Eq. (2), which measures the discrepancy between features learned for x2/x3/x4 SR after T-SNE transformation. In our latest version of paper, we have revised the motivation part for easier understanding.