GenSmoke-GS: A Multi-Stage Method for Novel View Synthesis from Smoke-Degraded Images Using a Generative Model

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cao, Qida, Hu, Xinyuan, Shi, Changyue, Ding, Jiajun, Yu, Zhou, Yu, Jun
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918429691215872
author Cao, Qida
Hu, Xinyuan
Shi, Changyue
Ding, Jiajun
Yu, Zhou
Yu, Jun
author_facet Cao, Qida
Hu, Xinyuan
Shi, Changyue
Ding, Jiajun
Yu, Zhou
Yu, Jun
contents This paper describes our method for Track 2 of the NTIRE 2026 3D Restoration and Reconstruction (3DRR) Challenge on smoke-degraded images. In this task, smoke reduces image visibility and weakens the cross-view consistency required by scene optimization and rendering. We address this problem with a multi-stage pipeline consisting of image restoration, dehazing, MLLM-based enhancement, 3DGS-MCMC optimization, and averaging over repeated runs. The main purpose of the pipeline is to improve visibility before rendering while limiting scene-content changes across input views. Experimental results on the challenge benchmark show improved quantitative performance and better visual quality than the provided baselines. The code is available at https://github.com/plbbl/GenSmoke-GS. Our method achieved a ranking of 1 out of 14 participants in Track 2 of the NTIRE 3DRR Challenge, as reported on the official competition website: https://www.codabench.org/competitions/13993/#/results-tab.
format Preprint
id arxiv_https___arxiv_org_abs_2604_03039
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle GenSmoke-GS: A Multi-Stage Method for Novel View Synthesis from Smoke-Degraded Images Using a Generative Model
Cao, Qida
Hu, Xinyuan
Shi, Changyue
Ding, Jiajun
Yu, Zhou
Yu, Jun
Computer Vision and Pattern Recognition
This paper describes our method for Track 2 of the NTIRE 2026 3D Restoration and Reconstruction (3DRR) Challenge on smoke-degraded images. In this task, smoke reduces image visibility and weakens the cross-view consistency required by scene optimization and rendering. We address this problem with a multi-stage pipeline consisting of image restoration, dehazing, MLLM-based enhancement, 3DGS-MCMC optimization, and averaging over repeated runs. The main purpose of the pipeline is to improve visibility before rendering while limiting scene-content changes across input views. Experimental results on the challenge benchmark show improved quantitative performance and better visual quality than the provided baselines. The code is available at https://github.com/plbbl/GenSmoke-GS. Our method achieved a ranking of 1 out of 14 participants in Track 2 of the NTIRE 3DRR Challenge, as reported on the official competition website: https://www.codabench.org/competitions/13993/#/results-tab.
title GenSmoke-GS: A Multi-Stage Method for Novel View Synthesis from Smoke-Degraded Images Using a Generative Model
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2604.03039