KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs | 100 lines of PyTorch code

Machine Learning: Implementation of the paper "KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs" in 100 lines of PyTorch code.
Novel view synthesis with NeRF at High FPS.
Udemy course: www.udemy.com/course/neural-r...
GitHub: github.com/MaximeVandegar/Pap...
-----------------------------------------------------------------------------------------------------
CONTACT: [e-mail]
#python #pytorch #nerf #neuralnetworks #machinelearning #artificialintelligence #deeplearning #data #unsupervisedlearning #research #neural #function #relu #positionalencoding #neuralrendering #rendering #neuralradiancefields #deeplearning #fastnerf #kilonerf #autoint #squeezenerf #3dreconstruction #novelviewsynthesis #instantngp #nvidia #radiance #fields #highfps #tiny #mlps

[@lgcondados]

A such an amazing content! Ty so much for that, I'm watching all your videos 😅

[@papersin100linesofcode]

Thank you so much for your comment!

[@tuanvu9755]

You're amazing, keep your great job!

[@papersin100linesofcode]

Thank you so much! :)

[@tuanvu9755]

@@papersin100linesofcode can I ask you some information? I'm wondering if your provided code will produce a similar-quality result compared to the NeRF code provided by the paper's authors?

[@papersin100linesofcode]

@@tuanvu9755 I am trying to get results as close as possible to the ones from the authors with most of my implementation. With NeRF, I usually do not compute the PSNR, and do not implement some components that would only improve results by a small margin, and make the code much more complicated. In short, running the official code should produce better results

[@gaurangtari7972]

Great work but how did you create the datasets?

[@papersin100linesofcode]

Thank you for your comment! I get asked this more and more, I will make a video about it in the coming weeks.