I am have the code from here. I have changed the images with some of my own and I have seen that SVM is not working. That is not a problem because I have found on another question the answer. Now I have changed the evaluation to NormalBayesClassifier. I am waiting for more than half an hour for the classifier to train. Is the NormalBayesClassifier::train function so slow? I have put 7 images in one class and 7 in the other (the size is about 640x480). (I have just 2 classes). I have changed the size of the images to much more smaller: around 180x100 and I have cropped the area where was too much background, but I am still waiting for training. What could be wrong?

Another thing: Is trainer working? I can do a lot of things meanwhile the training and my comp is not running slow...

  • $\begingroup$ Size of images doesnt matter. What is the size of the feature? $\endgroup$ – Aaron Jun 10 '14 at 18:44
  • $\begingroup$ Dictionary was 1000, when I have changed it to 250 it has got faster. The total number of features to classify was around 8000, but with 1000 words, it was more than 100000; this is the thing that makes it slow? $\endgroup$ – thedarkside ofthemoon Jun 11 '14 at 7:14
  • $\begingroup$ How to see the feature size? I have changed to GFTT. $\endgroup$ – thedarkside ofthemoon Jun 11 '14 at 7:22

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.