Change neural networks on web demo

[Avi Parshan]

I finally got the web demo to run properly! 

Next up, i want to use a new neural network instead of the reference CaffeNet, in order to shift

do I just need to change this class in app.py?

class ImagenetClassifier(object):
    default_args = {
        'model_def_file': (
            '{}/models/bvlc_reference_caffenet/deploy.prototxt'.format(REPO_DIRNAME)),
        'pretrained_model_file': (
            '{}/models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel'.format(REPO_DIRNAME)),
        'mean_file': (
            '{}/python/caffe/imagenet/ilsvrc_2012_mean.npy'.format(REPO_DIRNAME)),
        'class_labels_file': (
            '{}/data/ilsvrc12/synset_words.txt'.format(REPO_DIRNAME)),
        'bet_file': (
            '{}/data/ilsvrc12/imagenet.bet.pickle'.format(REPO_DIRNAME)),
    }
    for key, val in default_args.iteritems():
        if not os.path.exists(val):
            raise Exception(
                "File for {} is missing. Should be at: {}".format(key, val))
    default_args['image_dim'] = 256
    default_args['raw_scale'] = 255.

and import the new framework instead of this: 

import caffe