Open nightrome opened 7 years ago
Hi,
the VGG-16 model you referenced (here) requires that we subtract the mean image from the current image. Your code does not seem to do that, or? Have you checked what image classification performance you achieve with this model?
Hi,
the VGG-16 model you referenced (here) requires that we subtract the mean image from the current image. Your code does not seem to do that, or? Have you checked what image classification performance you achieve with this model?