Learning NeRF

This repository is initially created by Haotong Lin. Forked from LearningNeRF

TODO

• use opengl-based renderer to implement the depth peeling
• try other backbones of the K-plane encoder
• use the easyvolcap to visualize the 4D volume video

4K4D Result

test on single frame

after training for 400 iterations, the result is as follows: PSNR: 29.90dB

30 frames

after training for around 2k iterations, the result is as follows: