ELoG-GS: Dual-Branch Gaussian Splatting with Luminance-Guided Enhancement for Extreme Low-light 3D Reconstruction
ELoG-GS:双分支高斯点云生成与亮度引导增强用于极端低光3D重建
机构 * Shanghai Jiao Tong University(上海交通大学)
AI总结 本文提出ELoG-GS方法,结合学习点云初始化和亮度引导色彩增强,提升极端低光环境下3D重建的几何一致性和真实感。实验表明其在NTIRE 2026挑战赛中优于基线方法,达到更高的PSNR和SSIM。
Comments Our method achieved a ranking of 9 out of 148 participants in Track 1 of the NTIRE 3DRR Challenge, as reported on the official competition website: https://www.codabench.org/competitions/13854/