[CVPR2024] The official implementation of "MoCha-Stereo: Motif Channel Attention Network for Stereo Matching”. & [Arxiv] The official implementation of "Motif Channel Opened in a White-Box: Stereo Matching via Motif Correlation Graph"
MIT License
125
stars
4
forks
source link
Unable to find the implementation for reconstruction error #11
I'm trying to find the implementation of reconstruction error $E$ described in Eq. (7):
$$E=K_l(R-\frac{TN^T}D)K_r^{-1}I_r-I_l$$
but all i can find is a function called 'disp_wrap' on the line 220 of 'refinement.py':
ZYangChen
commented
3 months ago
I'm trying to find the implementation of reconstruction error $E$ described in Eq. (7): $$E=K_l(R-\frac{TN^T}D)K_r^{-1}I_r-I_l$$ but all i can find is a function called 'disp_wrap' on the line 220 of 'refinement.py':
looks very similar to Eq. (7), but without any configuration about camera intrinsic.