1) I have a list of 1.2 million ILSVRC2012 training/validation images with corresponding class codes, 000 to 999. Where can I find the name mapping for these codes?

2) Can anyone provide me a copy of the file "synset_words.txt" for the synset/name mapping of the ILSVRC2012 data set. Someone saw it mentioned on http://caffe.berkeleyvision.org/gathered/examples/imagenet.html but I can't find it anywhere on the web to include https://github.com/BVLC/caffe.

  • I now have the "sysnet_words" mapping for the 1000 classes of images for ILSVRC2012. Earlier, Minle wrote that I could resize my test image and input it into the CNN. If I convert my test JPG image to HKL format and resize it as needed, then how to I input this HKL test image into the CNN program alex_net.py? Please provide some suggestions based on this or other similar CNN programs. Also, what is the output vector variable called? As I understand, the softmax layer = 1000. How do I access the softmax layer output data? – arnold tunick Jun 17 '16 at 14:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.