when I try to run eval or render，i find the code will get some camera pose like "c2w，c2w_colmap" from poses_bounds.npy . So, what's mean of these parameter and poses_bounds.npy?
Another question is: In this papper, I can't understand how to obtain the camera poses of test views for rendering? Could you please explain in detail："Each test pose is initialised with the learned pose of the training frame that is closest to it".
when I try to run eval or render，i find the code will get some camera pose like "c2w，c2w_colmap" from poses_bounds.npy . So, what's mean of these parameter and poses_bounds.npy?
Another question is: In this papper, I can't understand how to obtain the camera poses of test views for rendering? Could you please explain in detail："Each test pose is initialised with the learned pose of the training frame that is closest to it".