Saved in:
Bibliographic Details
Main Authors: Shen, Licheng, Chow, Ho Ngai, Wang, Lingyun, Zhang, Tong, Wang, Mengqiu, Han, Yuxing
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.13694
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911884701073408
author Shen, Licheng
Chow, Ho Ngai
Wang, Lingyun
Zhang, Tong
Wang, Mengqiu
Han, Yuxing
author_facet Shen, Licheng
Chow, Ho Ngai
Wang, Lingyun
Zhang, Tong
Wang, Mengqiu
Han, Yuxing
contents Recent advancements in neural rendering techniques have significantly enhanced the fidelity of 3D reconstruction. Notably, the emergence of 3D Gaussian Splatting (3DGS) has marked a significant milestone by adopting a discrete scene representation, facilitating efficient training and real-time rendering. Several studies have successfully extended the real-time rendering capability of 3DGS to dynamic scenes. However, a challenge arises when training images are captured under vastly differing weather and lighting conditions. This scenario poses a challenge for 3DGS and its variants in achieving accurate reconstructions. Although NeRF-based methods (NeRF-W, CLNeRF) have shown promise in handling such challenging conditions, their computational demands hinder real-time rendering capabilities. In this paper, we present Gaussian Time Machine (GTM) which models the time-dependent attributes of Gaussian primitives with discrete time embedding vectors decoded by a lightweight Multi-Layer-Perceptron(MLP). By adjusting the opacity of Gaussian primitives, we can reconstruct visibility changes of objects. We further propose a decomposed color model for improved geometric consistency. GTM achieved state-of-the-art rendering fidelity on 3 datasets and is 100 times faster than NeRF-based counterparts in rendering. Moreover, GTM successfully disentangles the appearance changes and renders smooth appearance interpolation.
format Preprint
id arxiv_https___arxiv_org_abs_2405_13694
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Gaussian Time Machine: A Real-Time Rendering Methodology for Time-Variant Appearances
Shen, Licheng
Chow, Ho Ngai
Wang, Lingyun
Zhang, Tong
Wang, Mengqiu
Han, Yuxing
Computer Vision and Pattern Recognition
Recent advancements in neural rendering techniques have significantly enhanced the fidelity of 3D reconstruction. Notably, the emergence of 3D Gaussian Splatting (3DGS) has marked a significant milestone by adopting a discrete scene representation, facilitating efficient training and real-time rendering. Several studies have successfully extended the real-time rendering capability of 3DGS to dynamic scenes. However, a challenge arises when training images are captured under vastly differing weather and lighting conditions. This scenario poses a challenge for 3DGS and its variants in achieving accurate reconstructions. Although NeRF-based methods (NeRF-W, CLNeRF) have shown promise in handling such challenging conditions, their computational demands hinder real-time rendering capabilities. In this paper, we present Gaussian Time Machine (GTM) which models the time-dependent attributes of Gaussian primitives with discrete time embedding vectors decoded by a lightweight Multi-Layer-Perceptron(MLP). By adjusting the opacity of Gaussian primitives, we can reconstruct visibility changes of objects. We further propose a decomposed color model for improved geometric consistency. GTM achieved state-of-the-art rendering fidelity on 3 datasets and is 100 times faster than NeRF-based counterparts in rendering. Moreover, GTM successfully disentangles the appearance changes and renders smooth appearance interpolation.
title Gaussian Time Machine: A Real-Time Rendering Methodology for Time-Variant Appearances
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2405.13694