🐱 Cat vs Dogs 🐶

📝 About the project

Machine Learning model that receives an image and classifies it as cat or dog. The image dataset is available here.

Used tecnologies:

• Python
• Google Colab
• Pre-trained model
• TensorFlow

🌐 Machine Learning model working

🧠 Project Steps

Dataset

The first step is chose which dataset is gonna be used for make the machine learning model, as said before we use this dataset that contains 25,000 images from dogs and cats

Image processing

In our dataset we have several images but they are not equal and we need to create a pattern and make all images equal in dimensions, then we'll do it in code.

# Creating a directory for resized images
os.mkdir('/content/image resized')

# Resize all images for 224x224
original_folder = '/content/train/'
resized_folder = '/content/image resized/'

for i in range(2000):

  filename = os.listdir(original_folder)[i]
  img_path = original_folder+filename

  img = Image.open(img_path)
  img = img.resize((224, 224))
  img = img.convert('RGB')

  newImgPath = resized_folder+filename
  img.save(newImgPath)

Train/Test

Now we need do train our model with the resized images! but before this we need to say to our model what is cat and what is dog, for doing this we have to assume two labels cats = 0, dogs = 1

# Creating a for loop to assign labels
filenames = os.listdir('/content/image resized/')

labels = []

for i in range(2000):

  file_name = filenames[i]
  label = file_name[0:3]

  # dog = 1
  if label == 'dog':
    labels.append(1)
  # cat = 0
  else:
    labels.append(0)

Good! Now we gonna separate our dataset to train and test.

X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, random_state=2)

Finally we gonna effectively train the model. In this project we use mobilenet model who is a machine learning model we use a technique called 'Transfer Learning' that means take advantage of an existing model and use it in another scenario

mobilenet_model = 'https://tfhub.dev/google/tf2-preview/mobilenet_v2/feature_vector/4'
pretrained_model = hub.KerasLayer(mobilenet_model, input_shape=(224,224,3), trainable=False)

num_of_classes = 2
model = tf.keras.Sequential([
    pretrained_model,
    tf.keras.layers.Dense(num_of_classes)
])
model.summary()

model.compile(
    optimizer = 'adam',
    loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
    metrics = ['acc']
)

model.fit(X_train_scaled, Y_train, epochs=5)

🤔 How i use this project?

Kaggle Configuration

You need to have an account on Kaggle platform. Once logged in you have to go on account settings.

Click on 'Create New API Token' and you get 'kaggle.json' file.

Using .ipynb file

Download the 'dogs_and_cats_classification.ipynb' file on this github.

Open this file with Jupyter notebook or Google Colab!

Upload your 'kaggle.json' file on your project and run the code! Good! Now you can upload your dogs and cats images and ask to our model classify them!