Skip to content

Latest commit

 

History

History
12 lines (6 loc) · 757 Bytes

README.md

File metadata and controls

12 lines (6 loc) · 757 Bytes

NeRfFromScratch

Implementing NeRF from scratch, by following the original research paper.

I first started out by creating a 2D "nerf", which just memorizes a photo: https://github.com/shahanneda/NeRfFromScratch/blob/main/single-image.ipynb

And then I moved on to making it 3D and an actual nerf with volumetric rendering: https://github.com/shahanneda/NeRfFromScratch/blob/main/3d_image.ipynb

Then, to run overnight training runs, I extracted some code to classes and more proper python files, and also integrated with weights and biases to run many different training runs with different parameters: image