New SkyLume Dataset Tackles 3D Urban Reconstruction Under Varying Light

Researchers have introduced SkyLume, a large-scale aerial dataset designed to address challenges in 3D urban scene reconstruction under varying illumination conditions. The dataset comprises over 100,000 high-resolution UAV images captured across 10 urban regions at three different times of day to isolate illumination changes. It also includes per-scene LiDAR scans and ground-truth data for precise evaluation of geometry and appearance. Additionally, a new metric called the Temporal Consistency Coefficient (TCC) has been proposed to measure albedo stability and evaluate the disentanglement of light and material. AI

IMPACT This dataset aims to advance research in large-scale inverse rendering, geometry reconstruction, and novel view synthesis by providing a standardized resource for evaluating illumination-robust 3D reconstruction techniques.

RANK_REASON The cluster describes a new dataset and metric for 3D reconstruction research, published on arXiv. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• 3D Gaussian splatting
• lidar
• Neural Radiance Fields
• SkyLume
• Temporal Consistency Coefficient
• Zhuoxiao Li

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →