POSITIVE_IMAGE_DIR = os.path.join('data/Flickr_2800', args.pos_class) # let the user know what we're up to log.log("Training " + IDENTIFIER + " FROM " + POSITIVE_IMAGE_DIR) if args.transplant is not None: log.log("(Starting from " + args.transplant + "_t" + args.lesion_indicator + ")") else: log.log("(Starting from <blank net, random weights>)") # train it all with AlexNet() as alexnet: # Load data log.log("[Loading Data...]") all_data = Data() all_data.load_images(POSITIVE_IMAGE_DIR, POSITIVE_LABEL) all_data.load_images('data/Flickr_2800/notall', NEGATIVE_LABEL) train_data, test_data = all_data.split_train_test(train_split=0.90) # Initialize net if args.transplant is not None: load_transplant_dir = os.path.join(TRANSPLANT_DIR, args.transplant + '.ckpt') alexnet.load_transplant(load_transplant_dir) if args.lesion_indicator is not '': layers = [i for i, e in enumerate(args.lesion_indicator) if e == '0'] alexnet.lesion_layers(layers) # Train log.log("[Training...]") i = 0
from lib.alexnet import AlexNet from lib.data import Data from lib.perf import Perf import lib.utils as utils import matplotlib.pyplot as plt import lib.log as log import numpy as np import argparse import os # constants IMAGES_DIR = 'data/Flickr_2800' # actual run if __name__ == "__main__": # parse args out parser = argparse.ArgumentParser() parser.add_argument("img_class", type=str, help="The directory from Flickr_2800 to plot") parser.add_argument("x", type=int, help="Columns") parser.add_argument("y", type=int, help="Rows") args = parser.parse_args() d = Data() d.load_images(os.path.join(IMAGES_DIR, args.img_class), None) d.graph(args.x, args.y) plt.show()