低光下实现清晰3D场景重建，无需两阶段流程。
DelowlightSplat: Feed-Forward Gaussian Splatting for Lowlight 3D Scene Reconstruction
- 引入低光适配模块增强特征匹配能力
- 在低光条件下显著提升渲染质量与重建精度
- 适合机器人、AR/VR等低光环境应用
从稀疏姿态图像中进行新视角合成与3D重建是机器人和AR/VR的核心任务。然而，传统前馈式3D高斯重建在低光环境下因噪声、色彩偏移和对应不可靠而表现不佳。本文提出DelowlightSplat，一种面向低光的前馈式高斯点云渲染框架，可直接生成干净的新视角图像。我们构建了一个可控的多视图低光基准测试集，仅降级上下文视图而保持目标视图清晰。引入轻量级低光适配器进行残差增强以提升特征匹配性，并结合基于代价体的多视图推理，直接预测无噪3D高斯。实验表明，DelowlightSplat在低光条件下显著优于现有前馈方法及两阶段流水线。
原文摘要 · Abstract (English)
Novel-view synthesis and 3D reconstruction from sparse posed images are central to robotics and AR/VR. Yet, feed-forward 3D Gaussian reconstruction fails under lowlight due to noise, color shifts, and unreliable correspondence. We propose DelowlightSplat, a lowlight-aware feed-forward Gaussian splatting framework for clean novel-view rendering. We build a controllable multi-view lowlight benchmark by degrading only context views while keeping target views clean. We introduce a lightweight Lowlight Adapter for residual enhancement to improve matchability, and couple it with cost-volume-based multi-view inference to directly predict clean 3D Gaussians. Experiments show that DelowlightSplat significantly outperforms previous feed-forward method and two-stage pipeline under lowlight conditions.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。