Question about the bilinear model fitting

[xiaolin-mt]

Hi,

I read through the fitting code. It seems the released scripts only contain landmark alignment objectives. However, the paper describes that the pixel-pixel alignment objective and three other regularizations are applied. Could you please release the full optimization functions?

Thanks!