Computer Vision News - July 2020

2 Summary 16 We Tried For You Example Th e following example is extracted from the MONAI repository that can be found here , and, at this stage, it requires the second type of installation (through source code repo). GAN This example generates images that aremeant to resemble the hand X-rays provided as input, using a generative adversarial network (GAN). The MedNIST Dataset used is gathered from Bradley J. Erickson M.D., Ph.D. (Department of Radiology, Mayo Clinic ). The code below uses a combination of the torch library and Monai to load the data, construct the dataset, define the generator and discriminatory networks and finally perform the training. Networks and applied transforms are directly provided by Monai and, when used, this is highlighted in yellow. The last row is user-specific and should contain the path of the personal MONAI workspace. %matplotlib inline import os import sys import tarfile from urllib.request import urlopen from io import BytesIO import numpy as np import matplotlib.pyplot as plt from PIL import Image from IPython.display import clear_output from tqdm.notebook import tqdm sys.path.append(monai_directory)