How can I transform the poses (nerfstudio format) to opencv convention from OpenGL convention?
Opened this issue · 1 comments
thucz commented
I try to flip the y and z axes direction of the camera-to-world matrix but it seems that it is not enough. when I train generalizable Gaussian Splatting on DL3DV. The RGB loss suddenly increased a lot and failed.
c2w[..., :, 1:3] *= -1
ShengCN commented
It seems to be code dependent. Have you figured out what happened in your code?