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
Hello, thanks for great work and effort for HOISDF.
I have a question for model output.
After inference of the model(for one RGB image), i could get object rotation and translation results which are model_out["obj_rot_out"], model_out["obj_trans_out"]. From what I've checked, these results are (200,3) shape matrices. I observed these matrices and each 200 piece of length 3 vector looked similar. So, i use first vector among 200 vectors( euler angles rotation, and translation) for projecting mesh on the image, But it seems not working very well. Is there any specific format for these rotation and translation?
Thanks.
The text was updated successfully, but these errors were encountered:
Hi, have you solved this problem? I found problems when I did it. I think the mesh of the predicted object did not add mano_root. But the result of my code is that the object is tilted and does not correspond to the projection stolen on the RGB image, could you share your related code? That would be nice, thank you.
Hello, thanks for great work and effort for HOISDF.
I have a question for model output.
Thanks.
The text was updated successfully, but these errors were encountered: