SqueezeNeRF: Further factorized FastNeRF for memory-efficient inference

TL;DR

SqueezeNeRF introduces a highly memory-efficient neural radiance field model that maintains high-speed rendering, significantly reducing memory requirements compared to previous methods like FastNeRF, enabling practical real-time scene rendering.

Contribution

The paper proposes SqueezeNeRF, a novel factorized NeRF approach that drastically reduces memory usage while preserving high inference speed, improving real-world applicability.

Findings

SqueezeNeRF is over 60 times more memory-efficient than FastNeRF.

It achieves more than 190 frames per second during inference.

Maintains high-quality novel view synthesis with reduced resource demands.

Abstract

Neural Radiance Fields (NeRF) has emerged as the state-of-the-art method for novel view generation of complex scenes, but is very slow during inference. Recently, there have been multiple works on speeding up NeRF inference, but the state of the art methods for real-time NeRF inference rely on caching the neural network output, which occupies several giga-bytes of disk space that limits their real-world applicability. As caching the neural network of original NeRF network is not feasible, Garbin et al. proposed "FastNeRF" which factorizes the problem into 2 sub-networks - one which depends only on the 3D coordinate of a sample point and one which depends only on the 2D camera viewing direction. Although this factorization enables them to reduce the cache size and perform inference at over 200 frames per second, the memory overhead is still substantial. In this work, we propose SqueezeNeRF, which is more than 60 times memory-efficient than the sparse cache of FastNeRF and is still able to render at more than 190 frames per second on a high spec GPU during inference.