论文阅读:Dense Depth Priors for Neural Radiance Fields from Sparse Input Views
CVPR2022Preliminary首先我们由一组室内的RGB图像{Ii}i=0N−1,Ii∈[0,1]H×W×3\{I_i\}^{N-1}_{i=0},I_i\in[0,1]^{H\timesW\times3}{Ii}i=0N−1,Ii∈[0,1]H×W×3。通过SFM的方法,我们可以获得相机位姿pi∈R6p_i\in\mathbb{R}^6pi∈R6,内参矩阵Ki∈R3×3K_i\in\ma