Skip to content

Documentation.md

Latest commit

 

History

History
160 lines (129 loc) · 4.54 KB

Documentation.md

File metadata and controls

160 lines (129 loc) · 4.54 KB

Image Classification Model Documentation

Model Architecture

This model is a Convolutional Neural Network (CNN) designed for image classification on the CIFAR-10 dataset. The architecture consists of several convolutional layers followed by dense layers for classification.

Layers:

  1. Input Layer:

    • Input Shape: (32, 32, 3) for CIFAR-10 images.

  2. Convolutional Layers:

    • Conv2D Layer 1: 32 filters, kernel size (3, 3), ReLU activation.
    • MaxPooling2D Layer 1: Pool size (2, 2).
    • Conv2D Layer 2: 64 filters, kernel size (3, 3), ReLU activation.
    • MaxPooling2D Layer 2: Pool size (2, 2).
    • Conv2D Layer 3: 64 filters, kernel size (3, 3), ReLU activation.
    • MaxPooling2D Layer 3: Pool size (2, 2).

  3. Dense Layers:

    • Flatten Layer: Converts 3D feature maps to 1D feature vectors.
    • Dense Layer 1: 64 units, ReLU activation.
    • Output Layer: 10 units (one for each class), Softmax activation.

Model Summary:

model.summary()

Training Process

Data Preparation:

  1. Load CIFAR-10 Dataset:

    (train_images, train_labels), (test_images, test_labels) = tf.keras.datasets.cifar10.load_data()

  2. Normalize Images:

    train_images, test_images = train_images / 255.0, test_images / 255.0

  3. Split Training Data into Training and Validation Sets:

    from sklearn.model_selection import train_test_split
train_images, val_images, train_labels, val_labels = train_test_split(train_images, train_labels, test_size=0.2, random_state=42)

Data Augmentation:

  • Using ImageDataGenerator for augmenting the training data to improve generalization.
    datagen = ImageDataGenerator(
        rotation_range=40,
        width_shift_range=0.2,
        height_shift_range=0.2,
        shear_range=0.2,
        zoom_range=0.2,
        horizontal_flip=True,
        fill_mode='nearest'
    )

Model Compilation:

  • Compile the model with the Adam optimizer, sparse categorical cross-entropy loss, and accuracy as the metric.
model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])

Model Training:

  • Train the model with the augmented data for 30 epochs.
history = model.fit(
    datagen.flow(train_images, train_labels, batch_size=32),
    steps_per_epoch=len(train_images) // 32,
    epochs=30,
    validation_data=(val_images, val_labels)
)

Model Evaluation:

  • Evaluate the model on the test dataset to get the test accuracy.
test_loss, test_acc = model.evaluate(test_images, test_labels)
print(f"Test accuracy: {test_acc}")

Usage Instructions

Plot Training and Validation Accuracy/Loss:

Plotting the training and validation accuracy and loss over epochs.

acc = history.history['accuracy']
val_acc = history.history['val_accuracy']
loss = history.history['loss']
val_loss = history.history['val_loss']

epochs = range(len(acc))

plt.plot(epochs, acc, 'bo', label='Training acc')
plt.plot(epochs, val_acc, 'b', label='Validation acc')
plt.title('Training and validation accuracy')
plt.legend()

plt.figure()

plt.plot(epochs, loss, 'bo', label='Training loss')
plt.plot(epochs, val_loss, 'b', label='Validation loss')
plt.title('Training and validation loss')
plt.legend()

plt.show()

Save and Load the Model:

Save the trained model to a file.

model.save('cnn_image_classification_model.h5')

Load the saved model from the file.

model = tf.keras.models.load_model('cnn_image_classification_model.h5')

Predict New Images:

Functions to preprocess and predict new images.

def preprocess_image(image_path):
    img = tf.keras.utils.load_img(image_path, target_size=(32, 32))
    img_array = tf.keras.utils.img_to_array(img)
    img_array = np.expand_dims(img_array, axis=0) / 255.0
    return img_array

def predict_image(image_path):
    img_array = preprocess_image(image_path)
    predictions = model.predict(img_array)
    predicted_class = np.argmax(predictions, axis=1)
    return predicted_class[0]

Example Prediction:

Predict the class of a new image (uncomment the lines below to test).

# image_path = 'path_to_new_image.jpg'
# predicted_class = predict_image(image_path)
# print(f'Predicted class: {predicted_class}')

Example prediction on given images:

image_path = '/content/Vidwud_faceswap.jpg'
predicted_class = predict_image(image_path)
print(f'Predicted class: {predicted_class}')

image_path = '/content/gosau-3724039.jpg'
predicted_class = predict_image(image_path)
print(f'Predicted class: {predicted_class}')