Caffe Imagenet - test image example

[Zeevik]

hey guys,

I have a problem. when I train the Imagenet with my own Dataset i get accuracy about 0.6-0.7.

but when I run test on example image, like:

./build/examples/cpp_classification/classification.bin 
  models/bvlc_reference_caffenet/deploy.prototxt 
  models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel 
  data/ilsvrc12/imagenet_mean.binaryproto 
  data/ilsvrc12/synset_words.txt 
  examples/images/cat.jpg

I get always the same numbers of accuracy! (I have 2 outputs 0 & 1).

what am I doing wrong? my data set wrong? my synset_words.txt  wrong?

thanks for helping.

Zeev