Giters

lluisgomez / script_identification

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

I only got the accuarcy of 94.6%, not 94.8% on SIW-13

lulu1214 opened this issue · comments

commented

Dear lluisgomez, I only got the accuarcy of 94.6%, not 94.8% on SIW-13. I followd all steps according to your reference. Have you use some initial parameters for CNN? ?

commented

The results shown on the paper are obtained following the same steps, but due to stochastic GD and random sampling of the training set you may get a slightly different accuracy.