Adversarial_Observation package

Submodules

Adversarial_Observation.Attacks module

class Adversarial_Observation.Attacks.Config(epsilon=0.1, attack_method='fgsm')[source]: Bases: object

Adversarial_Observation.Attacks.compute_gradients(model, img, target_class)[source]

Adversarial_Observation.Attacks.fgsm_attack(input_batch_data: Tensor, model: Module, input_shape: tuple, epsilon: float = 0.0) → Tensor[source]

Adversarial_Observation.Attacks.generate_adversarial_examples(input_batch_data, model, method='fgsm', **kwargs)[source]

Adversarial_Observation.Attacks.log_metrics(success_rate, average_perturbation)[source]

Adversarial_Observation.Attacks.visualize_adversarial_examples(original, adversarial)[source]

Adversarial_Observation.utils module

Adversarial_Observation.utils.load_MNIST_data()[source]

Load the MNIST dataset and create data loaders.

Returns:: (train_loader, test_loader) - Data loaders for training and testing.
Return type:: tuple

Adversarial_Observation.utils.load_MNIST_model() → Sequential[source]

Build a convolutional neural network model.

Returns:: A convolutional neural network model for MNIST classification.
Return type:: nn.Sequential

Adversarial_Observation.utils.seed_everything(seed: int) → None[source]

Seed the random number generators for reproducibility.

Parameters:: seed (int) – The seed for random number generation.
Returns:: None

Adversarial_Observation.visualize module

Adversarial_Observation.visualize.visualize_gif(filenames: List[str], output_file: str = 'output.gif') → None[source]

Create a GIF from a list of image filenames.

Parameters:

filenames (List[str]) – List of image filenames.
output_file (str) – Output filename for the GIF (default: ‘output.gif’).

Returns:

None

Module contents