OccGaussian: 3D Gaussian Splatting for Occluded Human Rendering

TL;DR

OccGaussian introduces a fast, high-quality 3D human rendering method capable of handling occlusions in monocular videos, suitable for real-time applications, with significantly improved training and inference speeds.

Contribution

This work presents OccGaussian, a novel 3D Gaussian Splatting approach that drastically reduces training and rendering times while effectively handling occlusions in human rendering from monocular videos.

Findings

Achieves up to 160 FPS rendering with occluded inputs.

Reduces training time by 250x and inference time by 800x.

Performs comparably or better than state-of-the-art methods in occlusion scenarios.

Abstract

Rendering dynamic 3D human from monocular videos is crucial for various applications such as virtual reality and digital entertainment. Most methods assume the people is in an unobstructed scene, while various objects may cause the occlusion of body parts in real-life scenarios. Previous method utilizing NeRF for surface rendering to recover the occluded areas, but it requiring more than one day to train and several seconds to render, failing to meet the requirements of real-time interactive applications. To address these issues, we propose OccGaussian based on 3D Gaussian Splatting, which can be trained within 6 minutes and produces high-quality human renderings up to 160 FPS with occluded input. OccGaussian initializes 3D Gaussian distributions in the canonical space, and we perform occlusion feature query at occluded regions, the aggregated pixel-align feature is extracted to compensate for the missing information. Then we use Gaussian Feature MLP to further process the feature along with the occlusion-aware loss functions to better perceive the occluded area. Extensive experiments both in simulated and real-world occlusions, demonstrate that our method achieves comparable or even superior performance compared to the state-of-the-art method. And we improving training and inference speeds by 250x and 800x, respectively. Our code will be available for research purposes.