Physics-driven self-supervised learning for fast high-resolution robust 3D reconstruction of light-field microscopy

基于物理的自监督学习，用于快速、高分辨率、稳健的光场显微镜三维重建