It does not seem to use the mixed objective optimization？

[zhengge]

In train_val.prototxt， I do not find the mixed objective optimization.

layer {
 name: "euclidean-loss"
  type: "EuclideanLoss"
  bottom: "moon-fc"
  bottom: "labels"
  top: "loss"
}

Could you explain it? Thanks.

[bikong2]

i did not find the moon part, too.

[chaipangpang]

1.this project no relationship with the paper—“MOON”.
2.Why use the loss function —“EuclideanLoss”？It is better use “Sigmoid Cross-Entropy” or “HingeLoss”？

[BackT0TheFuture]

this repo does have noting to do with MOON loss.

@zhengge @bikong2 @chaipangpang

did you guys make MOON loss clear?
how to calculate src_dist ？

[klintqinami]

"Since CelebA has identical source and target distributions, we define the loss layer in (6) to weight all elements equally during backpropagation – which is equivalent to Euclidean loss between
the network output and the 40 binary attribute values."

[Page 9, MOON : A Mixed Objective Optimization Network for the Recognition of Facial Attributes]