Evaluating CrowdSplat: Perceived Level of Detail for Gaussian Crowds

TL;DR

This study evaluates how different levels of detail in Gaussian crowd avatars affect perceived realism, providing insights to optimize real-time crowd rendering in VR and gaming.

Contribution

It introduces a perceptual experiment assessing Gaussian avatar detail levels, informing LOD optimization for efficient, high-quality crowd rendering.

Findings

Perceived detail decreases with fewer Gaussians.

Viewing distance impacts perceived avatar detail.

Motion influences perceived realism.

Abstract

Efficient and realistic crowd rendering is an important element of many real-time graphics applications such as Virtual Reality (VR) and games. To this end, Levels of Detail (LOD) avatar representations such as polygonal meshes, image-based impostors, and point clouds have been proposed and evaluated. More recently, 3D Gaussian Splatting has been explored as a potential method for real-time crowd rendering. In this paper, we present a two-alternative forced choice (2AFC) experiment that aims to determine the perceived quality of 3D Gaussian avatars. Three factors were explored: Motion, LOD (i.e., #Gaussians), and the avatar height in Pixels (corresponding to the viewing distance). Participants viewed pairs of animated 3D Gaussian avatars and were tasked with choosing the most detailed one. Our findings can inform the optimization of LOD strategies in Gaussian-based crowd rendering, thereby helping to achieve efficient rendering while maintaining visual quality in real-time applications.