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

Hello, the published model model_150.ckpt is the best weighting?

yuwei66 opened this issue · comments

The measured accuracy does not meet the paper's specifications

Hi @q760019473, the results reported in the paper are obtained using a Matlab version of evaluation code. In this repo, we provide a Python version. Overall, the results are on par with the values in the paper with only slight differences in several scenarios.

Thank you ,got it