search

daviddmc / NeSVoR

NeSVoR is a package for GPU-accelerated slice-to-volume reconstruction.
MIT License
69 stars 16 forks source link

Visualizing learnt hash grid at different level #3

Closed alexpron closed 1 year ago

alexpron commented 1 year ago

Dear @daviddmc, I was trying to reproduce the visualization you made Fig.13 in the nesvor paper, i.e. the different levels of the learnt hashgrid. However I have not managed so far to obtain similar visualization. Is there some piece of code included in nesvor I could use to do that. If not, may you indicate me how to proceed ? So far I obtained patterns that are replicated, i.e. one brain then half a brain superimposed in the same figure, which seems weird. Best, Alex

daviddmc commented 1 year ago

Hi, seeing replicated brains is kind of expected due to the mod operation in the hash function. To get the results in the paper you would need to apply to brain mask to the hashgrid image.

alexpron commented 1 year ago

Hi @daviddmc thanks again for your answer. Actually we had an issue with tinycuda (described here https://github.com/NVlabs/tiny-cuda-nn/issues/286) which made the latent space uncorrect.