Cats and Dogs detection

CNN using Tensorflow 2 Beta

Dataset:

• https://www.kaggle.com/tongpython/cat-and-dog

Model:

tf.keras.Sequential( [
    tf.keras.layers.Conv2D(filters=32, kernel_size=(5,5), activation=tf.nn.leaky_relu, input_shape=(192,192,1)),
    tf.keras.layers.MaxPool2D(pool_size=(2,2), strides=(2,2)),
    tf.keras.layers.Conv2D(filters=32, kernel_size=(3,3), activation=tf.nn.leaky_relu),
    tf.keras.layers.MaxPool2D(pool_size=(2,2), strides=(2,2)),
    tf.keras.layers.Conv2D(filters=64, kernel_size=(3,3), activation=tf.nn.leaky_relu),
    tf.keras.layers.MaxPool2D(pool_size=(2,2), strides=(2,2)),
    tf.keras.layers.Conv2D(filters=128, kernel_size=(3,3), activation=tf.nn.leaky_relu),
    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(units=128, activation=tf.nn.leaky_relu),
    tf.keras.layers.Dense(units=64, activation=tf.nn.leaky_relu),
    tf.keras.layers.Dense(units=2, activation=tf.nn.softmax)
])