ChikaYan / d2nerf

Apache License 2.0
181 stars 14 forks source link

Thank you for your work, I have a question, why does this take so long to train, or the configuration required is so high. #17

Closed CHenYYff closed 10 months ago

ChikaYan commented 10 months ago

Hi,

Thanks for your interests. Our method is based on HyperNeRF, which is based on original NeRF and therefore is quite slow to train. There are some really cool fast NeRF implementation such as InstantNGP that allows the training to be significantly faster.