search

aadeshnpn / OSDN

Keras implementation for the research paper "Towards Open Set Deep Networks" A Bendale, T Boult, CVPR 2016
GNU Lesser General Public License v3.0
78 stars 21 forks source link

About the effect of the code #15

Open zhyhy opened 3 years ago

zhyhy commented 3 years ago

Hello, I tested your code on the MNIST data set and found that the effect is not so good. Can you achieve better open set recognition?

aadeshnpn commented 3 years ago

Thanks @zhyhy . Did you try training a different network on MNIST dataset and evaluate the performance? For me, the classification accuracy was around 97%.

xiangyangyu commented 3 years ago

I also encountered the same problem, using your code to run the performance is very poor, and I see your results in ipynb file are also not really good. Is there a problem somewhere?