kxhit
commented
1 year ago Hi again!
make_3D_grid is just for 3D mesh visualisation. The vectorised training part code is here, where we adopt functorch. A simple overview of using functorch to do vectorised training multi networks can be found here. Yeah, KiloNeRF is a very cool work and the idea of decomposition is similar. However, kilonerf splits the 3D space into uniform voxels, aims at faster rendering and needs customrized cuda kernels, while we decompose the scene into object-level, aim at flexible 3D reconstruction for robotics and the vectorised training is operated at Pytorch level.
Hi!!
I have questions regarding the vectorised training.....!!
As far as I understood, the vectorised training is somewhat similar to KiloNeRF, so I was wondering which part of the code is related to the vectorised training!
I am bit confused with the kiloNeRF since it looks like scene is uniformly distributed into n even blocks, so I was wondering whether the render_rays.make_3D_grid part is related to vectorised training....! If not, could you tell me which part of the code is related to vectorised training?
Thank you!!