Kilo Ne RF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs

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 MultiLayer Perceptron (MLP) millions of times, leading to slow rendering times, even on modern GPUs. In this paper, we demonstrate that significant speedups are 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 fastertoevaluate MLPs can be used. By combining this divideandconquer strategy with further optimizations, rendering is accelerated by two orders of magnitude compared to the original NeRF model without incurring high storage costs. Further, using teacherstudent distillation for training, we show that this speedup can be achieved without sacrificing visual quality.