inconsistent l_p value used in detector

[gehuangyi20]

In the original paper section 5.2.1, the paper said

Detector II and detector I(see Table 3) used the $L^2$ and $L^1$ norm to measure reconstruction error, respectively.

However, in the following implementation, you use L1 for detector II and L2 for detector I.

MagNet/test_defense.py

Line 14 in b115cf8

detector_I = AEDetector("./defensive_models/MNIST_I", p=2)

MagNet/test_defense.py

Line 15 in b115cf8

detector_II = AEDetector("./defensive_models/MNIST_II", p=1)

I am asking is this a just typo? Can you confirm which one should be the correct one.

It seems the code in the repo gives the better result.

[Trevillie]

Hi @gehuangyi20

As far as I recall, that should be a typo in the paper. Thanks for the notice.