Skip to main content

Examples

  • Image Classifier

Import Libraries​

import numpy as np
import random
import matplotlib.pyplot as plt
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Dense, Flatten

Download Dataset​

1000 images of cat and 1000 images of dog

pip install gdown
import gdown

files = {'labels.csv': 'https://drive.google.com/uc?id=1Zy3y6wBLZTI67BZhXzwQtgWJ8m50Oggl',
        'labels_test.csv': 'https://drive.google.com/uc?id=1cGDczlanBC59TbpNIe8_s0-yUtIrmnwm',
        'input.csv': 'https://drive.google.com/uc?id=1jQ16W4DJG1ZfJ_R7V_9cQIALwN3lruE5',
        'input_test.csv': 'https://drive.google.com/uc?id=1JHblwM88w4g70lZwiLDf6qPz8t0wZnd6',}


for output, url in files.items():
    print(url, output)
    gdown.download(url, output, quiet=False)

Load Dataset​

X_train = np.loadtxt('input.csv', delimiter = ',')
Y_train = np.loadtxt('labels.csv', delimiter = ',')

X_test = np.loadtxt('input_test.csv', delimiter = ',')
Y_test = np.loadtxt('labels_test.csv', delimiter = ',')

# reshape
X_train = X_train.reshape(len(X_train), 100, 100, 3)
Y_train = Y_train.reshape(len(Y_train), 1)

X_test = X_test.reshape(len(X_test), 100, 100, 3)
Y_test = Y_test.reshape(len(Y_test), 1)

# normalize
X_train = X_train/255.0
X_test = X_test/255.0

print("Shape of X_train: ", X_train.shape)
print("Shape of Y_train: ", Y_train.shape)
print("Shape of X_test: ", X_test.shape)
print("Shape of Y_test: ", Y_test.shape)
idx = random.randint(0, len(X_test))
plt.imshow(X_test[idx, :])
plt.show()

idx = random.randint(0, len(X_train))
plt.imshow(X_train[idx, :])
plt.show()

Create model​

model = Sequential([
    Conv2D(32, (3,3), activation = 'relu', input_shape = (100, 100, 3)),
    MaxPooling2D((2,2)),

    Conv2D(32, (3,3), activation = 'relu'),
    MaxPooling2D((2,2)),

    Flatten(),
    Dense(64, activation = 'relu'),
    Dense(1, activation = 'sigmoid')
])

or

model = Sequential()

model.add(Conv2D(32, (3,3), activation = 'relu', input_shape = (100, 100, 3)))
model.add(MaxPooling2D((2,2)))

model.add(Conv2D(32, (3,3), activation = 'relu'))
model.add(MaxPooling2D((2,2)))

model.add(Flatten())
model.add(Dense(64, activation = 'relu'))
model.add(Dense(1, activation = 'sigmoid'))

Compile model so that model can resume​

Adding cost function, and backpropogation

model.compile(loss = 'binary_crossentropy', optimizer = 'adam', metrics = ['accuracy'])

if need to use stochastic gradient descent instead of adam, this is used if we want to change the learning rate

opt = keras.optimizers.SGD(learning_rate = 0.001)
model.compile(loss = 'binary_crossentropy', optimizer = opt, metrics = ['accuracy'])

Train​

model.fit(X_train, Y_train, epochs = 5, batch_size = 64)

Evaluate​

model.evaluate(X_test, Y_test)

Inference / Making predictions​

idx2 = random.randint(0, len(Y_test))
plt.imshow(X_test[idx2, :])
plt.show()

y_pred = model.predict(X_test[idx2, :].reshape(1, 100, 100, 3))
cat_dog_pred = y_pred > 0.5

if(cat_dog_pred == 0):
    pred = 'dog'
else:
    pred = 'cat'

print("Our model says it is a :", pred, y_pred)

value y_pred gives a probabilistic value, close to 0 means it's 1st class i.e. dog, and close to 1 means it's 2nd class i.e. cat.

Cat vs Dog Classifier​