A new review paper published on arXiv addresses the challenges in underwater visual enhancement and 3D reconstruction. It details the physical models and surveys various methods, from traditional techniques to advanced data-driven approaches like Neural Radiance Fields and 3D Gaussian Splatting. The paper also includes quantitative and qualitative evaluations of current algorithms and outlines future research directions in underwater vision. AI
RANK_REASON This is a review paper on a specific topic within computer vision and AI. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →