search

sshan-zhao / GASDA

Geometry-Aware Symmetric Domain Adaptation for Monocular Depth Estimation, CVPR 2019
133 stars 23 forks source link

About Training Stage #12

Closed kingsleyluoxin closed 4 years ago

kingsleyluoxin commented 4 years ago

Hello, I am wondering what makes sense to train the f_s, f_t and gasda in different stages. Can that better optimize each model?

sshan-zhao commented 4 years ago

Hi, Sorry for replying to you late. Since I have given feedback via e-mail, I will close this issue, if you have questions, pls feel free to contact me or open a new issue.