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

arXiv cs.CV / 4/6/2026

📰 NewsSignals & Early TrendsModels & Research

共有:

Key Points

The paper presents GenSmoke-GS, a multi-stage pipeline for novel view synthesis using smoke-degraded images in the NTIRE 2026 3D Restoration and Reconstruction (3DRR) Challenge Track 2.
It improves cross-view rendering consistency under smoke by combining image restoration, dehazing, MLLM-based enhancement, and a 3DGS-MCMC optimization step followed by averaging over repeated runs.
The approach is designed to boost visibility prior to rendering while limiting unwanted changes to scene content across input views.
Experiments on the challenge benchmark show better quantitative metrics and improved visual quality versus the provided baselines.
GenSmoke-GS’s effectiveness is reflected by a top finish (1st of 14 participants) in the competition’s Track 2, and the authors provide code on GitHub.

Abstract

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.