search

hongsukchoi / 3DCrowdNet_RELEASE

Official Pytorch implementation of "Learning to Estimate Robust 3D Human Mesh from In-the-Wild Crowded Scenes", CVPR 2022
MIT License
155 stars 15 forks source link

Unstable test results #13

Open with-twilight opened 2 years ago

with-twilight commented 2 years ago

Hi, have you ever had a situation where the result of a retest is different from the result of the last test, and what may be the reason for this?

hongsukchoi commented 2 years ago

No.. did you change some codes?

with-twilight commented 2 years ago

Ok, I got it. Thank you.