mrezaei92/TriHorn-Net

The training result on the NYU dataset is 7.98, which does not achieve the effect in the paper.

whutdhz opened this issue · 1 comments

whutdhz commented

The training result on the NYU dataset is 7.98, which does not achieve the effect in the paper.

doduythao commented

The training result on the NYU dataset is 7.98, which does not achieve the effect in the paper.

@whutdhz have you get to 7.68 as the author claim ? can you share your configuration (aug also). Thank you