KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs

TL;DR

KiloNeRF introduces a method that uses thousands of small MLPs to significantly accelerate neural radiance field rendering, achieving real-time performance while maintaining high visual quality.

Contribution

The paper presents a novel divide-and-conquer approach with many tiny MLPs and distillation techniques to drastically speed up NeRF rendering without high storage costs.

Findings

Rendering speed increased by three orders of magnitude.

Maintained visual quality through teacher-student distillation.

Achieved real-time rendering on standard hardware.

Abstract

NeRF synthesizes novel views of a scene with unprecedented quality by fitting a neural radiance field to RGB images. However, NeRF requires querying a deep Multi-Layer Perceptron (MLP) millions of times, leading to slow rendering times, even on modern GPUs. In this paper, we demonstrate that real-time rendering is possible by utilizing thousands of tiny MLPs instead of one single large MLP. In our setting, each individual MLP only needs to represent parts of the scene, thus smaller and faster-to-evaluate MLPs can be used. By combining this divide-and-conquer strategy with further optimizations, rendering is accelerated by three orders of magnitude compared to the original NeRF model without incurring high storage costs. Further, using teacher-student distillation for training, we show that this speed-up can be achieved without sacrificing visual quality.