You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a couple of questions regarding 3DMM parameters and subsequently how you use them to project onto pixel space.
First I should mention that I am talking about demo_video.py
R and t_2d: you mentioned in the paper that first 12 elements of model output represent T in dimension of 3 by 4 including f, rotation matrix and translation matrix i.e., [t_2d, 0].T. The question has been raised in my mind that what the base frame is to which (coordinate sys) you are computing R and t_2d relative.
Moreover, if offset in recon_ver function indicates t_3d i.e., [t_2d, 0].T, so we expect the third element of vector should be zero but I checked that and the result denies the fact mentioned above.
projection: when you try to plot using cv_draw_landmarks, you just plot first two elements of ver (that is in 3D) without any projections and using intrinsics.
Please let me know if you have any explanation for these concerns since I am really interested in this pipeline.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Hi there,
I have a couple of questions regarding 3DMM parameters and subsequently how you use them to project onto pixel space.
First I should mention that I am talking about demo_video.py
Moreover, if offset in recon_ver function indicates t_3d i.e., [t_2d, 0].T, so we expect the third element of vector should be zero but I checked that and the result denies the fact mentioned above.
Please let me know if you have any explanation for these concerns since I am really interested in this pipeline.
Thanks.
Beta Was this translation helpful? Give feedback.
All reactions