def main(): #print("hello world") for luck in_arg = train_input_args() print(" data directory =",in_arg.data_dir, "\n save directory =", in_arg.save_dir, "\n model architecture =", in_arg.arch, "\n hidden units =", in_arg.hidden_units, "\n learning rate =", in_arg.learning_rate, "\n epochs =", in_arg.epochs, "\n device =", in_arg.device) image_data, data_loader = load_images() model, optimizer, criterion = build_model(in_arg.arch , in_arg.hidden_units, in_arg.learning_rate) model = train_model(model,data_loader, in_arg.epochs, criterion, optimizer, in_arg.device) save_model(model, optimizer, in_arg.arch, in_arg.data_dir, in_arg.save_dir,image_data) print("-"*40)
import numpy as np import functions as func # Download images X = func.load_images("train-images-idx3-ubyte.gz") # Input number try: val = int(input('Enter a number 0 or more to 9999 or less: ')) image = X[val] if(val > 9999): raise IndexError except IndexError: # IndexError print('Invalid index.') raise except: # Error print('Invalid string.') raise # Run task y = func.forward(image) num = np.argmax(y) print(num)
#Volleyball simulation import pygame as pg, math import functions as fn pg.init() #screen size screen_size=(800,600) #loads and transforms images bg,rod,playerr,playerb,volleyball,overlay1,overlay2=fn.load_images() #loads fonts TNR1=pg.font.SysFont("Times New Roman", 50) TNR2=pg.font.SysFont("Times New Roman", 30) PETC=TNR1.render('Press enter to exit...',1,(0,0,0)) paused=TNR1.render('PAUSED',1,(0,0,0)) welcome=TNR1.render('WELCOME!',1,(0,0,0)) choice=TNR2.render('Select the difficult level:',1,(0,0,0)) choice1=TNR2.render('Easy - Press 1',1,(0,0,0)) choice2=TNR2.render('Medium - Press 2',1,(0,0,0)) choice3=TNR2.render('Difficult - Press 3',1,(0,0,0)) #initializes object for framerate clock=pg.time.Clock() #making screen screen=pg.display.set_mode(screen_size) pg.display.set_caption('Volleyball')
import numpy as np import cv2 from matplotlib import pyplot as plt from functions import load_images from functions import filter_image from functions import analyze_derivative from functions import get_half_width_on_half_hight # Load images original_images, images_names = load_images("*.bmp") assert len(original_images) > 0 print("found " + str(len(original_images)) + " images") for i, image, image_name in zip(range(len(original_images)), original_images, images_names): print('processing', i, 'image') fil_im = filter_image(image, vetrical_alignment=30, horizontal_alignment=30) img_der = analyze_derivative(fil_im, axis=1, name=image_name[:-4] + "_derivative") temp_slice = img_der[1000, :].copy() widths = np.zeros(4) env_size = 50 for i in range(4): ind_max, widths[i] = get_half_width_on_half_hight(temp_slice, env_size)
from keras.models import load_model from functions import SAVED_MODEL_PATH, TRAINING_PATH, EVALUATE_PATH, get_latest_model, load_images ######################################################################################################################## model_name = get_latest_model() model = load_model(model_name) print('-' * 80) print("Loaded model " + model_name) # model.summary() ######################################################################################################################## print("Loading evaluation dataset ...") X_test, y_test = load_images(EVALUATE_PATH, 68, 69) print("... done!") print(X_test.shape, y_test.shape) y_predict = model.predict(X_test)[0] y_predict = cv2.normalize(y_predict, None, alpha=0, beta=255, norm_type=cv2.NORM_MINMAX, dtype=cv2.CV_8U) mask = cv2.normalize(y_test[0], None, alpha=0, beta=255,
from keras.models import load_model from functions import SAVED_MODEL_PATH, TRAINING_PATH, EVALUATE_PATH, get_latest_model, load_images ######################################################################################################################## # Training parameters batch_size = 16 epochs = 6 TIME = 3 ######################################################################################################################## print("Loading training dataset ...") X_train, y_train = load_images(TRAINING_PATH, 0, 10000) # X_train, y_train = load_images(TRAINING_PATH, 5000 * TIME, 5000 * (TIME + 1)) print("... done!") print(X_train.shape, y_train.shape) print("Loading evaluation dataset ...") X_test, y_test = load_images(EVALUATE_PATH, 0, 500) print("... done!") print(X_test.shape, y_test.shape) ######################################################################################################################## model_name = get_latest_model() model = load_model(model_name) print('-' * 80) print("Loaded model " + model_name)
from keras.applications import vgg16, inception_v3, resnet50, mobilenet tensorflow_master = "" checkpoint_path = "NIPS/inception-v3/inception_v3.ckpt" input_dir = "NIPS/images" max_epsilon = 16.0 image_width = 299 image_height = 299 batch_size = 1000 eps = 2.0 * max_epsilon / 255.0 batch_shape = [batch_size, image_height, image_width, 3] num_classes = 1001 categories = pd.read_csv("NIPS/categories.csv") image_classes = pd.read_csv("NIPS/images.csv") image_iterator = load_images(input_dir, batch_shape) # get first batch of images filenames, images = next(image_iterator) image_metadata = pd.DataFrame({ "ImageId": [f[:-4] for f in filenames] }).merge(image_classes, on="ImageId") true_classes = image_metadata["TrueLabel"].tolist() target_classes = true_labels = image_metadata["TargetClass"].tolist() #true_classes_names = (pd.DataFrame({"CategoryId": true_classes}) # .merge(categories, on="CategoryId")["CategoryName"].tolist()) true_classes_names = [] for i in true_classes: