I would like to start by thanking you for creating legible MPL repo. Your work has been invaluable in advancing the field of computer vision, and I am truly grateful for your contributions.
I am writing to ask you one question about dot_product in your code:
But in official implementation of MPL, h is computed by:
which means h = L_S_new - L_S_old
And this is my writing of h:
Seem like we have the same idea. But why is it different from author's ?
If you figured it out, please inform me. Thank you so much.
Dear ifsheldon,
I would like to start by thanking you for creating legible MPL repo. Your work has been invaluable in advancing the field of computer vision, and I am truly grateful for your contributions. I am writing to ask you one question about dot_product in your code:
But in official implementation of MPL, h is computed by:
which means h = L_S_new - L_S_old
And this is my writing of h:
Seem like we have the same idea. But why is it different from author's ?
If you figured it out, please inform me. Thank you so much.