BEAT - pkorshunov/gmm-spoofing/4

This algorithm is a legacy one. The API has changed since its implementation. New versions and forks will need to be updated.

Endpoint Groups 1

Algorithms have at least one input and one output. All algorithm endpoints are organized in groups. Groups are used by the platform to indicate which inputs and outputs are synchronized together. The first group is automatically synchronized with the channel defined by the block in which the algorithm is deployed.

Group: main

Endpoint Name	Data Format	Nature
features	system/array_2d_floats/1	Input
class	system/text/1	Input
classifier	pkorshunov/two-classes-gmm/1	Output

Parameters 2

Parameters allow users to change the configuration of an algorithm when scheduling an experiment

Name	Description	Type	Default	Range/Choices
number-of-gaussians	The number of Gaussian Components	uint32	100
maximum-number-of-iterations	The maximum number of iterations for the EM algorithm	uint32	10

import numpy
import bob.learn.em

class Algorithm:

def __init__(self):
        self.number_of_gaussians = 100
        self.max_iterations = 10
        self.positives = []
        self.negatives = []

def setup(self, parameters):
        self.number_of_gaussians = parameters.get('number-of-gaussians',
                                                  self.number_of_gaussians)

self.max_iterations = parameters.get('maximum-number-of-iterations',
                                             self.max_iterations)

return True

def create_gmm(self, training_data):
        input_size = training_data.shape[1]
        # create the KMeans and GMM machine
        kmeans = bob.learn.em.KMeansMachine(int(self.number_of_gaussians), input_size)
        gmm = bob.learn.em.GMMMachine(int(self.number_of_gaussians), input_size)

# create the KMeansTrainer
        kmeans_trainer = bob.learn.em.KMeansTrainer('RANDOM_NO_DUPLICATE')

# train using the KMeansTrainer
        bob.learn.em.train(kmeans_trainer, kmeans, training_data, int(self.max_iterations))

(variances, weights) = kmeans.get_variances_and_weights_for_each_cluster(training_data)
        means = kmeans.means

# initialize the GMM
        gmm.means = means
        gmm.variances = variances
        gmm.weights = weights

# train the GMM
        trainer = bob.learn.em.ML_GMMTrainer()
        bob.learn.em.train(trainer, gmm, training_data, int(self.max_iterations))
        
        return gmm

def process(self, inputs, outputs):
        # accumulates the input data in different
        # containers for hit or miss
        feature_vector = inputs["features"].data.value
        if inputs["class"].data.text == 'real':
            self.positives.append(feature_vector)
        else:
            self.negatives.append(feature_vector)

if not(inputs.hasMoreData()):
            # create array set used for training
            self.positives = numpy.vstack(self.positives)
            self.negatives = numpy.vstack(self.negatives)

gmm_real = self.create_gmm(self.positives)
            gmm_attacks = self.create_gmm(self.negatives)

# outputs data
            outputs["classifier"].write({
                "model_one": {
                        'weights':              gmm_real.weights,
                        'means':                gmm_real.means,
                        'variances':            gmm_real.variances,
                        'variance_thresholds':  gmm_real.variance_thresholds,
                 },
                "model_two": {
                        'weights':              gmm_attacks.weights,
                        'means':                gmm_attacks.means,
                        'variances':            gmm_attacks.variances,
                        'variance_thresholds':  gmm_attacks.variance_thresholds,
                 },
            })

return True

The code for this algorithm in Python
The ruler at 80 columns indicate suggested POSIX line breaks (for readability).
The editor will automatically enlarge to accomodate the entirety of your input
Use keyboard shortcuts for search/replace and faster editing. For example, use Ctrl-F (PC) or Cmd-F (Mac) to search through this box

Implements a GMM-based training, each GMM model for each of two types of data

Experiments

Attestation:

Privacy:

Status:

			Updated	Name	Databases/Protocols	Analyzers
				pkorshunov/pkorshunov/isv-asv-pad-fusion-complete/1/asv_isv-pad_gmm-fusion_lr-pa	avspoof/2@physicalaccess_verify_train,avspoof/2@physicalaccess_verification,avspoof/2@physicalaccess_verification_spoof,avspoof/2@physicalaccess_verify_train_spoof,avspoof/2@physicalaccess_antispoofing	pkorshunov/spoof-score-fusion-roc_hist/1
				pkorshunov/pkorshunov/speech-pad-simple/1/speech-pad_gmm-pa	avspoof/2@physicalaccess_antispoofing	pkorshunov/simple_antispoofing_analyzer/4

This table shows the number of times this algorithm has been successfully run using the given environment. Note this does not provide sufficient information to evaluate if the algorithm will run when submitted to different conditions.

algorithms pkorshunov gmm-spoofing 4

Endpoint Groups 1

Group: main

Parameters 2

Experiments

algorithms

pkorshunov

gmm-spoofing

4