Irving Yu

这段数据集制作代码我看到了。在计算损失的时候，由于 cx 和 cy 是固定的，周围那一圈赋值的数据其实没有用到。我的理解没问题吧。

Hi, @chuzcjoe Sorry to reply to you so late, I've been busy doing something else lately... In the `AFLW2000` landmarks, some points have a minimum value less than 0 and...

@wqz960 I don’t seem to encounter this problem. Have you cleaned the data set correctly?

@wqz960 In the AFLW2000 landmarks, some points have a minimum value less than 0 and a maximum value greater than width or height. You need to clip them.

When I use `tesnorflow2.0`, I set the `BatchSize` to `128`, although the `nan` will appear, the model still recycles the face. This is really amazing. ToT I suspect it may...

That's because when you stack three multi-scale features together, you get a new tensor which shape is `(bs, 176)`. Please see the [code](https://github.com/polarisZhao/PFLD-pytorch/blob/master/models/pfld.py#L129) in `models/pfld.py` @sainisanjay @Sridhar98