zhang-tao-whu/e2ec

Coarse -> refine process

Opened this issue · 0 comments

gasking commented

Q1： Is it possible to effectively fit the regression points to the real points in discrete spatial dimensions with a simple l1 loss?