BEAT - tpereira/dct_features

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

This algorithm is splittable

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
image	system/array_2d_floats/1	Input
features	system/array_2d_floats/1	Output

Parameters 5

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

Name	Description	Type	Default
block-overlap	Number of pixels to overlap (in both directions)	uint32	11
number-of-components	Number of components kept. Is n-1 since the DC component is excluded	uint32	45
norm-dcts	For a set of blocks, will normalize the coefficients	bool	True
block-size	Size of the block in pixels (in both directions)	uint32	12
norm-blocks	Will normalize the pixels in the block before the DCT computation	bool	True

import numpy
import bob.ip.base

class Algorithm:

def __init__(self):
        self.block_size = 12
        self.block_overlap = 11
        self.number_of_components = 45
        self.norm_blocks = True
        self.norm_dcts = True

def setup(self, parameters):
        self.block_size = parameters.get('block-size', self.block_size)
        self.block_overlap = parameters.get('block-overlap', self.block_overlap)
        self.number_of_components = parameters.get('number-of-components',
                                                   self.number_of_components)
        self.norm_blocks = parameters.get('norm-blocks', self.norm_blocks)
        self.norm_dcts = parameters.get('norm-dcts', self.norm_dcts)

self.extractor = bob.ip.base.DCTFeatures(self.number_of_components, 
                                                   (self.block_size, self.block_size), 
                                                   (self.block_overlap, self.block_overlap),
                                                   bool(self.norm_blocks), 
                                                   bool(self.norm_dcts)
                                                 )

return True

def process(self, inputs, outputs):
        float_image = inputs["image"].data.value

outputs["features"].write({
            'value': numpy.vstack(self.extractor(float_image))
        })

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

Extract DCT features for the parts-based Face Recognition System described in [McCool2009], [McCool2013].

This algorithm relies on the Bob library.

The input, image, is a two-dimensional array of floats (64 bits) corresponding to one image. The outputs, features, is a two-dimensional array of floats (64 bits) corresponding to the DCT coeficients of each block.

[McCool2009]

McCool, Christopher, and Sébastien Marcel: Parts-based face verification using local frequency bands. Advances in Biometrics. Springer Berlin Heidelberg, 2009. 259-268.

[McCool2013]

McCool, Christopher, et al. &quot;Session variability modelling for face authentication.&quot; IET biometrics 2.3 (2013): 117-129.

Experiments

Attestation:

Privacy:

Status:

			Updated	Name	Databases/Protocols	Analyzers
				tpereira/tpereira/full_isv/2/btas2015_periocular_mobio-female_det_bobv2-0	mobio/1@female	tutorial/eerhter_postperf_iso/1

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 tpereira dct_features_bobv2-0 1

Endpoint Groups 1

Group: main

Parameters 5

Experiments

algorithms

tpereira

dct_features_bobv2-0

1