Would you like to share the initial depth estimation methods?

[cdowen]

Hello, would you like to give out the code for the initial estimation of depth map? It can help us evaluate the method on other datasets (for example, the KITTI dataset). Thank you greatly.

[abelguima]

Yes, If it's possible, for us it will be better, and we can understand what you have been done in your project. Thanks in advance. All the best. Abel

On Wed, Jun 12, 2019 at 11:17 PM cdowen notifications@github.com wrote:

Hello, would you like to give out the code for the initial estimation of depth map? It can help us evaluate the method on other datasets (for example, the KITTI dataset). Thank you greatly.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/google/mannequinchallenge/issues/4?email_source=notifications&email_token=ABISC3DFO3PCQKVKLY6TIW3P2G33BA5CNFSM4HXVRFH2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GZGTYGQ, or mute the thread https://github.com/notifications/unsubscribe-auth/ABISC3E3WYSCPSBD74IPDQ3P2G33BANCNFSM4HXVRFHQ .

[fcole]

It's unfortunately not easy to make a robust release of the depth estimation code, since it involves running COLMAP as well as optical flow estimation and combining the results (as described in the paper). If we get enough interest we may revisit this issue, though.

[cdowen]

Thanks for your work and the depth estimation results are impressive. So would you like to share just the depth estimation method (the plane-plus-parallax method) during the evaluation? It should not be too troublesome but necessary for evaluation on other datasets. And it could help us greatly. Thank you.