How to predict multiple images in Keras at a time using multiple-processing (e.g. with different CPUs)?

Question

I have a lot of PNG images that I want to classify, using a trained CNN model.

To speed up the process, I would like to use multiple-processing with CPUs (I have 72 avai

Answer 1

One python package I know that may help you is joblib. Hope it can solve your problem.

from joblib import Parallel, delayed

# load model
mymodel = load_model('190704_1_fcs_plotclassifier.h5')

# Define callback function to collect the output in 'outcomes'
outcomes = []

def collect_result(result):
    global outcomes
    outcomes.append(result)

# Define prediction function

def prediction(img):
    img = cv2.resize(img,(49,49))
    img = img.astype('float32') / 255
    img = np.reshape(img,[1,49,49,3])       

    status = mymodel.predict(img)
    status = status[0][1]

    return(status)

# Define evaluate function

def evaluate(i,figure):

    # predict the propability of the picture to be in class 0 or 1
    img = cv2.imread(figure)
    status = prediction(img)

    outcome = [figure, status]
    return(i,outcome)

outcomes = Parallel(n_jobs=72)(delayed(evaluate)(i,figure) for figure in listoffigurepaths)