Skip to content
Snippets Groups Projects
Commit bbe0e08c authored by Guillaume HEUSCH's avatar Guillaume HEUSCH
Browse files

[preprocessor, extractor] fixed docstrings (again)

parent ba80ebba
No related branches found
No related tags found
1 merge request!53WIP: rPPG as features for PAD
Hide whitespace changes
Inline Side-by-side
bob/pad/face/extractor/LTSS.py
+ 20
− 26
View file @ bbe0e08c
......@@ -15,29 +15,23 @@ class LTSS(Extractor, object):
The features are described in the following article:
@Article {
Author = {Muckenhirn, Hannah and Korshunov, Pavel and
Magimai-Doss, Mathew and Marcel, Sebastien }
Title = {Long-Term Spectral Statistics for Voice Presentation
Attack Detection},
Journal = {IEEE/ACM Trans. Audio, Speech and Lang. Proc.},
Volume = {25},
Number = {11},
Pages = {2098--2111},
year = 2017
}
H. Muckenhirn, P. Korshunov, M. Magimai-Doss, and S. Marcel
Long-Term Spectral Statistics for Voice Presentation Attack Detection,
IEEE/ACM Trans. Audio, Speech and Language Processing. vol 25, n. 11, 2017
Attributes
----------
framerate: int
window_size : :obj:`int`
The size of the window where FFT is computed
framerate : :obj:`int`
The sampling frequency of the signal (i.e the framerate ...)
nfft: int
nfft : :obj:`int`
Number of points to compute the FFT
debug: bool
debug : :obj:`bool`
Plot stuff
concat: bool
concat : :obj:`bool`
Flag if you would like to concatenate features from the 3 color channels
time: int
time : :obj:`int`
The length of the signal to consider (in seconds)
"""
......@@ -46,19 +40,19 @@ class LTSS(Extractor, object):
Parameters
----------
window_size: int
window_size : :obj:`int`
The size of the window where FFT is computed
framerate: int
framerate : :obj:`int`
The sampling frequency of the signal (i.e the framerate ...)
nfft: int
nfft : :obj:`int`
Number of points to compute the FFT
concat: bool
Flag if you would like to concatenate features from the 3 color channels
debug: bool
debug : :obj:`bool`
Plot stuff
time: int
concat : :obj:`bool`
Flag if you would like to concatenate features from the 3 color channels
time : :obj:`int`
The length of the signal to consider (in seconds)
"""
super(LTSS, self).__init__()
self.framerate = framerate
......@@ -73,12 +67,12 @@ class LTSS(Extractor, object):
Parameters
----------
signal: numpy.ndarray
signal : :py:class:`numpy.ndarray`
The signal
Returns
-------
ltss: numpy.ndarray
:py:class:`numpy.ndarray`
The spectral statistics of the signal.
"""
......
bob/pad/face/extractor/LiFeatures.py
+ 15
− 24
View file @ bbe0e08c
......@@ -13,43 +13,34 @@ class LiFeatures(Extractor, object):
"""Compute features from pulse signals in the three color channels.
The features are described in the following article:
@InProceedings{li-icpr-2016,
Author = {Li, X. and Komulainen, J. and Zhao, G. and Yuen, P-C.
and Pietik\"ainen, M.},
Title = {Generalized {F}ace {A}nti-spoofing by {D}etecting
{P}ulse {F}rom {F}ace {V}ideos},
BookTitle = {Intl {C}onf. on {P}attern {R}ecognition ({ICPR})},
Pages = {4244-4249},
year = 2016
}
X. Li, J. Komulainen, G. Zhao, P-C Yuen and M. Pietikainen,
Generalized Face Anti-spoofing by Detecting Pulse From Face Videos
Intl Conf. on Pattern Recognition (ICPR), 2016.
Attributes
----------
framerate: int
framerate : :obj:`int`
The sampling frequency of the signal (i.e the framerate ...)
nfft: int
nfft : :obj:`int`
Number of points to compute the FFT
debug: bool
debug : :obj:`bool`
Plot stuff
"""
def __init__(self, framerate=25, nfft=512, debug=False, **kwargs):
"""Init function
Parameters
----------
framerate: int
framerate : :obj:`int`
The sampling frequency of the signal (i.e the framerate ...)
nfft: int
nfft : :obj:`int`
Number of points to compute the FFT
debug: bool
debug : :obj:`bool`
Plot stuff
"""
super(LiFeatures, self).__init__()
self.framerate = framerate
self.nfft = nfft
......@@ -61,13 +52,13 @@ class LiFeatures(Extractor, object):
Parameters
----------
signal: numpy.ndarray
signal : :py:class:`numpy.ndarray`
The signal
Returns
-------
feature: numpy.ndarray
the computed features
-------
:py:class:`numpy.ndarray`
the computed features
"""
# sanity check
......
bob/pad/face/extractor/PPGSecure.py
+ 11
− 21
View file @ bbe0e08c
......@@ -15,27 +15,17 @@ class PPGSecure(Extractor, object):
The feature are extracted according to what is described in
the following article:
@InProceedings{nowara-afgr-2017,
Author = {E. M. Nowara and A. Sabharwal and A. Veeraraghavan},
Title = {P{PGS}ecure: {B}iometric {P}resentation {A}ttack
{D}etection {U}sing {P}hotopletysmograms},
BookTitle = {I{EEE} {I}ntl {C}onf on {A}utomatic {F}ace and
{G}esture {R}ecognition ({AFGR})},
Volume = {},
Number = {},
Pages = {56-62},
issn = {},
seq-number = {69},
year = 2017
}
E.M Nowara, A. Sabharwal and A. Veeraraghavan,
"PPGSecure: Biometric Presentation Attack Detection using Photoplethysmograms",
IEEE Intl Conf. on Automatic Face and Gesture Recognition, 2017.
Attributes
----------
framerate: int
framerate : :obj:`int`
The sampling frequency of the signal (i.e the framerate ...)
nfft: int
nfft : :obj:`int`
Number of points to compute the FFT
debug: bool
debug : :obj:`bool`
Plot stuff
"""
......@@ -44,11 +34,11 @@ class PPGSecure(Extractor, object):
Parameters
----------
framerate: int
framerate : :obj:`int`
The sampling frequency of the signal (i.e the framerate ...)
nfft: int
nfft : :obj:`int`
Number of points to compute the FFT
debug: bool
debug : :obj:`bool`
Plot stuff
"""
......@@ -63,12 +53,12 @@ class PPGSecure(Extractor, object):
Parameters
----------
signal: numpy.ndarray
signal : :py:class:`numpy.ndarray`
The signal
Returns
-------
fft: numpy.ndarray
:py:class:`numpy.ndarray`
the computed FFT features
"""
......
bob/pad/face/preprocessor/Chrom.py
+ 19
− 21
View file @ bbe0e08c
......@@ -25,26 +25,26 @@ class Chrom(Preprocessor, object):
The pulse is extracted according to the CHROM algorithm.
See the documentation of :py:mod:`bob.rppg.base`
See the documentation of `bob.rppg.base`
Attributes
----------
skin_threshold: float
skin_threshold : :obj:`float`
The threshold for skin color probability
skin_init: bool
skin_init : :obj:`bool`
If you want to re-initailize the skin color distribution at each frame
framerate: int
framerate : :obj:`int`
The framerate of the video sequence.
bp_order: int
bp_order : :obj:`int`
The order of the bandpass filter
window_size: int
window_size : :obj:`int`
The size of the window in the overlap-add procedure.
motion: float
motion : :obj:`float`
The percentage of frames you want to select where the
signal is "stable". 0 mean all the sequence.
debug: boolean
debug : :obj:`bool`
Plot some stuff
skin_filter: :py:class:`bob.ip.skincolorfilter.SkinColorFilter`
skin_filter : :py:class:`bob.ip.skincolorfilter.SkinColorFilter`
The skin color filter
"""
......@@ -54,24 +54,23 @@ class Chrom(Preprocessor, object):
Parameters
----------
skin_threshold: float
skin_threshold : :obj:`float`
The threshold for skin color probability
skin_init: bool
skin_init : :obj:`bool`
If you want to re-initailize the skin color distribution at each frame
framerate: int
framerate : :obj:`int`
The framerate of the video sequence.
bp_order: int
bp_order : :obj:`int`
The order of the bandpass filter
window_size: int
window_size : :obj:`int`
The size of the window in the overlap-add procedure.
motion: float
motion : :obj:`float`
The percentage of frames you want to select where the
signal is "stable". 0 mean all the sequence.
debug: boolean
debug : :obj:`bool`
Plot some stuff
"""
super(Chrom, self).__init__()
self.skin_threshold = skin_threshold
self.skin_init = skin_init
......@@ -87,14 +86,14 @@ class Chrom(Preprocessor, object):
Parameters
----------
frames: :py:class:`bob.bio.video.utils.FrameContainer`
frames : :py:class:`bob.bio.video.utils.FrameContainer`
video data
annotations: :py:class:`dict`
annotations : :py:class:`dict`
the face bounding box, as follows: ``{'topleft': (row, col), 'bottomright': (row, col)}``
Returns
-------
pulse: numpy.ndarray
:obj:`numpy.ndarray`
The pulse signal
"""
......@@ -239,4 +238,3 @@ class Chrom(Preprocessor, object):
pyplot.show()
return pulse
bob/pad/face/preprocessor/Li.py
+ 14
− 15
View file @ bbe0e08c
......@@ -22,7 +22,7 @@ class Li(Preprocessor):
The pulse is extracted according to Li's CVPR 14 algorithm.
See the documentation of :py:mod:`bob.rppg.base`
See the documentation of `bob.rppg.base`
Note that this is a simplified version of the original
pulse extraction algorithms (mask detection in each
......@@ -31,17 +31,17 @@ class Li(Preprocessor):
Attributes
----------
indent: int
indent : :obj:`int`
Indent (in percent of the face width) to apply to keypoints to get the mask.
lamda_: int
lamda_ : :obj:`int`
the lamba value of the detrend filter
window: int
window : :obj:`int`
The size of the window of the average filter
framerate: int
framerate : :obj:`int`
The framerate of the video sequence.
bp_order: int
bp_order : :obj:`int`
The order of the bandpass filter
debug: bool
debug : :obj:`bool`
Plot some stuff
"""
......@@ -51,19 +51,18 @@ class Li(Preprocessor):
Parameters
----------
indent: int
indent : :obj:`int`
Indent (in percent of the face width) to apply to keypoints to get the mask.
lamda_: int
lamda_ : :obj:`int`
the lamba value of the detrend filter
window: int
window : :obj:`int`
The size of the window of the average filter
framerate: int
framerate : :obj:`int`
The framerate of the video sequence.
bp_order: int
bp_order : :obj:`int`
The order of the bandpass filter
debug: bool
debug : :obj:`bool`
Plot some stuff
"""
super(Li, self).__init__(**kwargs)
......@@ -86,7 +85,7 @@ class Li(Preprocessor):
Returns
-------
pulse: numpy.ndarray
:obj:`numpy.ndarray`
The pulse signal, in each color channel (RGB)
"""
......
bob/pad/face/preprocessor/PPGSecure.py
+ 10
− 21
View file @ bbe0e08c
......@@ -22,41 +22,30 @@ class PPGSecure(Preprocessor):
The pulse is extracted according to what is described in
the following article:
@InProceedings{nowara-afgr-2017,
Author = {E. M. Nowara and A. Sabharwal and A. Veeraraghavan},
Title = {P{PGS}ecure: {B}iometric {P}resentation {A}ttack
{D}etection {U}sing {P}hotopletysmograms},
BookTitle = {I{EEE} {I}ntl {C}onf on {A}utomatic {F}ace and
{G}esture {R}ecognition ({AFGR})},
Volume = {},
Number = {},
Pages = {56-62},
issn = {},
seq-number = {69},
year = 2017
}
E.M Nowara, A. Sabharwal and A. Veeraraghavan,
"PPGSecure: Biometric Presentation Attack Detection using Photoplethysmograms",
IEEE Intl Conf. on Automatic Face and Gesture Recognition, 2017.
Attributes
----------
framerate: int
framerate : :obj:`int`
The framerate of the video sequence.
bp_order: int
bp_order : :obj:`int`
The order of the bandpass filter
debug: bool
debug : :obj:`bool`
Plot some stuff
"""
def __init__(self, framerate=25, bp_order=32, debug=False, **kwargs):
"""Init function
Parameters
----------
framerate: int
framerate : :obj:`int`
The framerate of the video sequence.
bp_order: int
bp_order : :obj:`int`
The order of the bandpass filter
debug: bool
debug : :obj:`bool`
Plot some stuff
"""
......@@ -83,7 +72,7 @@ class PPGSecure(Preprocessor):
Returns
-------
pulse: numpy.ndarray
:obj:`numpy.ndarray`
The pulse signal, in each color channel (RGB)
"""
......
bob/pad/face/preprocessor/SSR.py
+ 2
− 2
View file @ bbe0e08c
......@@ -23,7 +23,7 @@ class SSR(Preprocessor, object):
The pulse is extracted according to the SSR algorithm.
See the documentation of :py:mod:`bob.rppg.base`
See the documentation of :py:module::`bob.rppg.base`
Attributes
----------
......@@ -35,7 +35,7 @@ class SSR(Preprocessor, object):
The temporal stride.
debug: boolean
Plot some stuff
skin_filter: :py:class:`bob.ip.skincolorfilter.SkinColorFilter`
skin_filter: :py:class::`bob.ip.skincolorfilter.SkinColorFilter`
The skin color filter
"""
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment