image-recognizer/main.py

from ui.app import App

if __name__ == '__main__':
    app = App()
    app.mainloop()

# import numpy as np
# from matplotlib import pyplot as plt
#
# import matplotlib
#
# matplotlib.use("TkAgg")
# np.random.seed(0)
#
# from utils.mnist import MNISTNeuralNet
#
# # Set the precision to 3 decimal places
# np.set_printoptions(precision=8, suppress=True)
#
# from utils.load_mnist import get_test_images, get_test_labels, get_train_images, get_train_labels
#
# train_images = get_train_images()
# train_labels = get_train_labels()
#
# mnist_neural_net = MNISTNeuralNet()
# losses = mnist_neural_net.train(train_images, train_labels, 0.0001, 100)

# test_images = get_test_images()
# test_labels = get_test_labels()
# results = mnist_neural_net.forward(test_images)
# predictions = results.argmax(axis=1)
#
# correct = predictions == test_labels
# incorrect = predictions != test_labels
# accuracy = mnist_neural_net.accuracy(results, test_labels)
# # Create figure and axes
# fig, ax = plt.subplots(figsize=(10, 5))
#
# ax.hist(test_labels[correct], bins=np.arange(11)-0.5, alpha=0.5, label="Correct", color="green")
# ax.hist(test_labels[incorrect], bins=np.arange(11)-0.5, alpha=0.5, label="Incorrect", color="red")
# ax.set_xticks(range(10))
# ax.set_xlabel("True Label")
# ax.set_ylabel("Count")
# ax.set_title(f"Accuracy {accuracy}")
# ax.legend()
#
# fig.show()

# while True:
#     plt.pause(0.1)

####################
##   Draw image   ##
####################
# Create a figure and axes
# fig, ax = plt.subplots()

# Initial matrix displayed
# initial_data = np.array(images[0])
# mat = ax.matshow(initial_data.reshape(28, 28), cmap='bwr')
# fig.show()

# Redraw the canvas
# fig.canvas.draw()
# fig.canvas.flush_events()
#
# plt.pause(20)