Event-guided 3D Gaussian Splatting for Dynamic Human and Scene Reconstruction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yin, Xiaoting, Shi, Hao, Yang, Kailun, Zhai, Jiajun, Guo, Shangwei, Wang, Lin, Wang, Kaiwei
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909802088628224
author Yin, Xiaoting
Shi, Hao
Yang, Kailun
Zhai, Jiajun
Guo, Shangwei
Wang, Lin
Wang, Kaiwei
author_facet Yin, Xiaoting
Shi, Hao
Yang, Kailun
Zhai, Jiajun
Guo, Shangwei
Wang, Lin
Wang, Kaiwei
contents Reconstructing dynamic humans together with static scenes from monocular videos remains difficult, especially under fast motion, where RGB frames suffer from motion blur. Event cameras exhibit distinct advantages, e.g., microsecond temporal resolution, making them a superior sensing choice for dynamic human reconstruction. Accordingly, we present a novel event-guided human-scene reconstruction framework that jointly models human and scene from a single monocular event camera via 3D Gaussian Splatting. Specifically, a unified set of 3D Gaussians carries a learnable semantic attribute; only Gaussians classified as human undergo deformation for animation, while scene Gaussians stay static. To combat blur, we propose an event-guided loss that matches simulated brightness changes between consecutive renderings with the event stream, improving local fidelity in fast-moving regions. Our approach removes the need for external human masks and simplifies managing separate Gaussian sets. On two benchmark datasets, ZJU-MoCap-Blur and MMHPSD-Blur, it delivers state-of-the-art human-scene reconstruction, with notable gains over strong baselines in PSNR/SSIM and reduced LPIPS, especially for high-speed subjects.
format Preprint
id arxiv_https___arxiv_org_abs_2509_18566
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Event-guided 3D Gaussian Splatting for Dynamic Human and Scene Reconstruction
Yin, Xiaoting
Shi, Hao
Yang, Kailun
Zhai, Jiajun
Guo, Shangwei
Wang, Lin
Wang, Kaiwei
Computer Vision and Pattern Recognition
Robotics
Image and Video Processing
Reconstructing dynamic humans together with static scenes from monocular videos remains difficult, especially under fast motion, where RGB frames suffer from motion blur. Event cameras exhibit distinct advantages, e.g., microsecond temporal resolution, making them a superior sensing choice for dynamic human reconstruction. Accordingly, we present a novel event-guided human-scene reconstruction framework that jointly models human and scene from a single monocular event camera via 3D Gaussian Splatting. Specifically, a unified set of 3D Gaussians carries a learnable semantic attribute; only Gaussians classified as human undergo deformation for animation, while scene Gaussians stay static. To combat blur, we propose an event-guided loss that matches simulated brightness changes between consecutive renderings with the event stream, improving local fidelity in fast-moving regions. Our approach removes the need for external human masks and simplifies managing separate Gaussian sets. On two benchmark datasets, ZJU-MoCap-Blur and MMHPSD-Blur, it delivers state-of-the-art human-scene reconstruction, with notable gains over strong baselines in PSNR/SSIM and reduced LPIPS, especially for high-speed subjects.
title Event-guided 3D Gaussian Splatting for Dynamic Human and Scene Reconstruction
topic Computer Vision and Pattern Recognition
Robotics
Image and Video Processing
url https://arxiv.org/abs/2509.18566