Using the pretrained model get AUC=80.1310% on ped2

[AndersonZhangyq]

I used the provided pretrained model, but got AUC=80.1310% on ped2, which is much lower than the AUC reported in the paper(Ours w DPU: 96.9) I moved the pretrained model (i.e., ped2_model_1000.pth) to folder exp/ped2/log/, and run the command as

python Test.py --model_dir /home/zyq/code/MPN/exp/ped2/log/ped2_model_1000.pth

Is there anything wrong with my procedure? How can I reproduce the result reported in the paper?

[abhishekaich27]

I observed similar issue for Shanghai-Tech as well. Although I trained the model by myself.

[devashishlohani]

I also got the AUC=80.13% with the pre-trained model on Ped2 dataset in the unsupervised setting (Test.py).
However, after training the model on Ped2, I retested after 1000 epochs. The result is 96.13%. Note, I used alpha (weight for feature anomaly score) as 0.6. Hope it helps!

[1449362433]

How do you reproduce it?