/ngp_jax

Implementation of "instant neural graphics primitives (instant ngp)" with JAX

Primary LanguageJupyter Notebook

Instant Neural Graphics Primitives in JAX (non-official)

Table of Contents

Overview

The instant ngp in pure JAX still needs optimization. Currently, it is way slower than the CUDA version.

Prerequisites

  • Set up a virtual Python3 environment, then install necessary packages with this command 
    conda create --name jax_ngp python=3
conda activate jax_ngp
pip install -r requirements.txt
  • Or, just open Google Colab and run notebook_train.ipynb

Reference

[1] Mildenhall, Ben, et al. "NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis." ECCV (2020).

[2] Müller, Thomas, et al. "Instant Neural Graphics Primitives with a Multiresolution Hash Encoding." SIGGRAPH (2022).