Using trained caffe model in python script, added value scaling and mean.

prediction.py

import sys
import caffe
import cv2
import Image
import numpy as np
from scipy.misc import imresize

caffe_root = "/home/simon/Workspaces/caffe/"

#MODEL_FILE = caffe_root + 'models/placesCNN/places205CNN_deploy.prototxt'
#PRETRAINED = caffe_root + 'models/placesCNN/places205CNN_iter_300000.caffemodel'

MODEL_FILE = caffe_root + 'models/customAlex/deploy.prototxt'
PRETRAINED = caffe_root + 'models/customAlex/caffe_alexnet_train_iter_1500.caffemodel'

net = caffe.Net(MODEL_FILE, PRETRAINED, caffe.TEST)
caffe.set_mode_cpu()

blob = caffe.proto.caffe_pb2.BlobProto()
data = open(caffe_root + 'models/customAlex/mean_train.binaryproto' , 'rb' ).read()
blob.ParseFromString(data)

meanArray = np.array( caffe.io.blobproto_to_array(blob) ).transpose(3,2,1,0)
meanArray = meanArray[:,:,:,0]

img = caffe.io.load_image(caffe_root + 'examples/images/Thess-11516his_110178,03.jpg')
img = img - meanArray
img = imresize(img, [227, 227])
img = img.astype(np.uint8)

imageData = np.asarray([img.transpose(2, 1, 0)])
imageData = np.divide(imageData, 255.0)

# out = net.forward_all(data=np.asarray([img.transpose(2, 0, 1)]))
out = net.forward_all(data=imageData)
print("Predicted class is #{}.".format(out['prob'][0].argmax()))

#imagenet_labels_filename = caffe_root + 'models/placesCNN/categoryIndex_places205.csv'
imagenet_labels_filename = caffe_root + 'models/customAlex/indexLabelMapping.txt'

labels = np.loadtxt(imagenet_labels_filename, str, delimiter='\s')

top_k = net.blobs['prob'].data[0].flatten().argsort()[-1: -6: -1]
print out['prob'][0]
print top_k
print labels[top_k]