Gaussian Wave Splatting for Computer-Generated Holography

TL;DR

This paper introduces Gaussian Wave Splatting, a novel algorithm that converts neural scene representations into photorealistic holograms supporting occlusions and view-dependent effects, advancing holographic display technology.

Contribution

It presents a closed-form Gaussian-to-hologram transform and an efficient Fourier domain approximation, enabling real-time, accurate hologram generation from neural scene data.

Findings

Supports accurate occlusions and view-dependent effects

Enables efficient parallelizable hologram computation

Integrates neural rendering with holographic displays

Abstract

State-of-the-art neural rendering methods optimize Gaussian scene representations from a few photographs for novel-view synthesis. Building on these representations, we develop an efficient algorithm, dubbed Gaussian Wave Splatting, to turn these Gaussians into holograms. Unlike existing computer-generated holography (CGH) algorithms, Gaussian Wave Splatting supports accurate occlusions and view-dependent effects for photorealistic scenes by leveraging recent advances in neural rendering. Specifically, we derive a closed-form solution for a 2D Gaussian-to-hologram transform that supports occlusions and alpha blending. Inspired by classic computer graphics techniques, we also derive an efficient approximation of the aforementioned process in the Fourier domain that is easily parallelizable and implement it using custom CUDA kernels. By integrating emerging neural rendering pipelines with holographic display technology, our Gaussian-based CGH framework paves the way for next-generation holographic displays.