Hi Zhang,
as you said in the paper, 'a sub-network m(·) learns to produce an inlier probability map or mask, highlighting the content in the feature maps that contribute much for the homography estimation. ' I know clearly what is the function of this predictor but could you pls show me in detail what is the sub-network like and how does it work to reject the outlier? Thank you.
Hi Zhang, as you said in the paper, 'a sub-network m(·) learns to produce an inlier probability map or mask, highlighting the content in the feature maps that contribute much for the homography estimation. ' I know clearly what is the function of this predictor but could you pls show me in detail what is the sub-network like and how does it work to reject the outlier? Thank you.