diff --git a/bob/bio/base/config/examples/gabor_mobio-male.py b/bob/bio/base/config/examples/gabor_mobio-male.py
index a20453ed28611697c0c063162fbab8f54f74ed92..8336c510b5cbd927647abd607fb6339f99bcbcc7 100644
--- a/bob/bio/base/config/examples/gabor_mobio-male.py
+++ b/bob/bio/base/config/examples/gabor_mobio-male.py
@@ -26,6 +26,7 @@ database = DatabaseConnector(
protocol="mobile0-male",
)
)
+database.allow_score_multiple_references = True
# Using face crop
CROPPED_IMAGE_HEIGHT = 80
@@ -74,9 +75,8 @@ algorithm = AlgorithmAsBioAlg(callable=gabor_jet, features_dir=base_dir, allow_s
#algorithm = AlgorithmAsBioAlg(callable=gabor_jet, features_dir=base_dir)
-# comment out the code below to disable dask
-from bob.pipelines.mixins import estimator_dask_it, mix_me_up
-from bob.bio.base.pipelines.vanilla_biometrics.mixins import BioAlgDaskMixin
+from bob.bio.base.pipelines.vanilla_biometrics import VanillaBiometrics, dask_vanilla_biometrics
+
+#pipeline = VanillaBiometrics(transformer, algorithm)
+pipeline = dask_vanilla_biometrics(VanillaBiometrics(transformer, algorithm), npartitions=48)
-transformer = estimator_dask_it(transformer)
-algorithm = mix_me_up([BioAlgDaskMixin], algorithm)
diff --git a/bob/bio/base/config/examples/lda_atnt_legacy.py b/bob/bio/base/config/examples/lda_atnt_legacy.py
index 4ad30b856f56079a5037b029c1f7f2ef9152be22..dd6f93edc46fb041870d2dea18ab66f02fd5257a 100644
--- a/bob/bio/base/config/examples/lda_atnt_legacy.py
+++ b/bob/bio/base/config/examples/lda_atnt_legacy.py
@@ -59,11 +59,7 @@ algorithm = CheckpointDistance(features_dir="./example/")
# algorithm = Distance()
-# comment out the code below to disable dask
-from bob.pipelines.mixins import estimator_dask_it, mix_me_up
-from bob.bio.base.pipelines.vanilla_biometrics.mixins import (
- BioAlgDaskMixin,
-)
+from bob.bio.base.pipelines.vanilla_biometrics import VanillaBiometrics, dask_vanilla_biometrics
-transformer = estimator_dask_it(transformer)
-algorithm = mix_me_up([BioAlgDaskMixin], algorithm)
+#pipeline = VanillaBiometrics(transformer, algorithm)
+pipeline = dask_vanilla_biometrics(VanillaBiometrics(transformer, algorithm))
diff --git a/bob/bio/base/config/examples/lda_atnt_legacy_all_legacy.py b/bob/bio/base/config/examples/lda_atnt_legacy_all_legacy.py
index 6bedb447b6efd21de7be6022a81ed633a8874e38..d9736430ea82c0623137ee53ae3bf6919e40242e 100644
--- a/bob/bio/base/config/examples/lda_atnt_legacy_all_legacy.py
+++ b/bob/bio/base/config/examples/lda_atnt_legacy_all_legacy.py
@@ -58,10 +58,7 @@ transformer = make_pipeline(
algorithm = AlgorithmAsBioAlg(callable=lda, features_dir="./example/")
-from bob.pipelines.mixins import estimator_dask_it, mix_me_up
-from bob.bio.base.pipelines.vanilla_biometrics.mixins import (
- BioAlgDaskMixin,
-)
+from bob.bio.base.pipelines.vanilla_biometrics import VanillaBiometrics, dask_vanilla_biometrics
-transformer = estimator_dask_it(transformer)
-algorithm = mix_me_up([BioAlgDaskMixin], algorithm)
+#pipeline = VanillaBiometrics(transformer, algorithm)
+pipeline = dask_vanilla_biometrics(VanillaBiometrics(transformer, algorithm))
diff --git a/bob/bio/base/config/examples/pca_atnt.py b/bob/bio/base/config/examples/pca_atnt.py
index 3cd159b1d621d98a71e25e8b76d2a03d9ef05330..e75daf2752badb3930f34a10c62943455fa38970 100644
--- a/bob/bio/base/config/examples/pca_atnt.py
+++ b/bob/bio/base/config/examples/pca_atnt.py
@@ -11,6 +11,7 @@ import os
base_dir = "example"
database = DatabaseConnector(AtntBioDatabase(original_directory="./atnt", protocol="Default"))
+database.allow_scoring_with_all_biometric_references = True
transformer = make_pipeline(
CheckpointSampleLinearize(features_dir=os.path.join(base_dir, "linearize")),
@@ -26,6 +27,7 @@ from bob.bio.base.pipelines.vanilla_biometrics.mixins import (
BioAlgDaskMixin,
)
-transformer = estimator_dask_it(transformer)
-algorithm = mix_me_up([BioAlgDaskMixin], algorithm)
+from bob.bio.base.pipelines.vanilla_biometrics import VanillaBiometrics, dask_vanilla_biometrics
+pipeline = VanillaBiometrics(transformer, algorithm)
+#pipeline = dask_vanilla_biometrics(VanillaBiometrics(transformer, algorithm))
diff --git a/bob/bio/base/config/examples/pca_mobio-male.py b/bob/bio/base/config/examples/pca_mobio-male.py
index 45fc2049b45226045d17db9de5f5c500eb454b6c..668a1e6f7cbdc34dbae1f1ac5bcf40299db07bcf 100644
--- a/bob/bio/base/config/examples/pca_mobio-male.py
+++ b/bob/bio/base/config/examples/pca_mobio-male.py
@@ -44,12 +44,7 @@ transformer = make_pipeline(
)
algorithm = CheckpointDistance(features_dir="./example/")
-# comment out the code below to disable dask
-from bob.pipelines.mixins import estimator_dask_it, mix_me_up
-from bob.bio.base.pipelines.vanilla_biometrics.mixins import (
- BioAlgDaskMixin,
-)
-
-transformer = estimator_dask_it(transformer)
-algorithm = mix_me_up([BioAlgDaskMixin], algorithm)
+from bob.bio.base.pipelines.vanilla_biometrics import VanillaBiometrics, dask_vanilla_biometrics
+#pipeline = VanillaBiometrics(transformer, algorithm)
+pipeline = dask_vanilla_biometrics(VanillaBiometrics(transformer, algorithm), npartitions=48)
diff --git a/bob/bio/base/pipelines/vanilla_biometrics/__init__.py b/bob/bio/base/pipelines/vanilla_biometrics/__init__.py
index edbb4090fca046b19d22d3982711084621bff3be..d67835735a076de0842a1f9467eaf54eb53a4fb7 100644
--- a/bob/bio/base/pipelines/vanilla_biometrics/__init__.py
+++ b/bob/bio/base/pipelines/vanilla_biometrics/__init__.py
@@ -1,4 +1,6 @@
# see https://docs.python.org/3/library/pkgutil.html
from pkgutil import extend_path
+from .pipeline import VanillaBiometrics, dask_vanilla_biometrics
+
__path__ = extend_path(__path__, __name__)
diff --git a/bob/bio/base/pipelines/vanilla_biometrics/pipeline.py b/bob/bio/base/pipelines/vanilla_biometrics/pipeline.py
index 4df6aa8975f834925f17496e7a5ec6e552a72af7..c5d0cbf487c708d2b8d6a8e5ef01cfecbc8fa897 100644
--- a/bob/bio/base/pipelines/vanilla_biometrics/pipeline.py
+++ b/bob/bio/base/pipelines/vanilla_biometrics/pipeline.py
@@ -14,89 +14,160 @@ import numpy
logger = logging.getLogger(__name__)
-def biometric_pipeline(
- background_model_samples,
- biometric_reference_samples,
- probe_samples,
- transformer,
- biometric_algorithm,
- allow_scoring_with_all_biometric_references=False,
-):
- logger.info(
- f" >> Vanilla Biometrics: Training background model with pipeline {transformer}"
- )
-
- # Training background model (fit will return even if samples is ``None``,
- # in which case we suppose the algorithm is not trainable in any way)
- transformer = train_background_model(background_model_samples, transformer)
-
- logger.info(
- f" >> Creating biometric references with the biometric algorithm {biometric_algorithm}"
- )
-
- # Create biometric samples
- biometric_references = create_biometric_reference(
- biometric_reference_samples, transformer, biometric_algorithm
- )
-
- logger.info(
- f" >> Computing scores with the biometric algorithm {biometric_algorithm}"
- )
-
- # Scores all probes
- return compute_scores(
+class VanillaBiometrics(object):
+ """
+ Vanilla Biometrics Pipeline
+
+ This is the backbone of most biometric recognition systems.
+ It implements three subpipelines and they are the following:
+
+ - :py:class:`VanillaBiometrics.train_background_model`: Initializes or trains your transformer.
+ It will run :py:meth:`sklearn.base.BaseEstimator.fit`
+
+ - :py:class:`VanillaBiometrics.create_biometric_reference`: Creates biometric references
+ It will run :py:meth:`sklearn.base.BaseEstimator.transform` followed by a sequence of
+ :py:meth:`bob.bio.base.pipelines.vanilla_biometrics.abstract_classes.BioAlgorithm.enroll`
+
+ - :py:class:`VanillaBiometrics.compute_scores`: Computes scores
+ It will run :py:meth:`sklearn.base.BaseEstimator.transform` followed by a sequence of
+ :py:meth:`bob.bio.base.pipelines.vanilla_biometrics.abstract_classes.BioAlgorithm.score`
+
+
+ Example
+ -------
+ >>> from sklearn.pipeline import make_pipeline
+ >>> from bob.bio.base.pipelines.vanilla_biometrics.implemented import Distance
+ >>> transformer = make_pipeline(estimator_1, estimator_2)
+ >>> biometric_algoritm = Distance()
+ >>> pipeline = VanillaBiometrics(transformer, biometric_algoritm)
+ >>> pipeline(samples_for_training_back_ground_model, samplesets_for_enroll, samplesets_for_scoring)
+
+
+ To run this pipeline using Dask, used the function :py:func:`dask_vanilla_biometrics`.
+
+ Example
+ -------
+ >>> pipeline = VanillaBiometrics(transformer, biometric_algoritm)
+ >>> pipeline = dask_vanilla_biometrics(pipeline)
+ >>> pipeline(samples_for_training_back_ground_model, samplesets_for_enroll, samplesets_for_scoring).compute()
+
+
+ Parameters:
+ -----------
+
+ transformer: :py:class`sklearn.pipeline.Pipeline` or a `sklearn.base.BaseEstimator`
+ Transformer that will preprocess your data
+
+ biometric_algorithm: :py:class:`bob.bio.base.pipelines.vanilla_biometrics.abstract_classes.BioAlgorithm`
+ Biometrics algorithm object that implements the methods `enroll` and `score` methods
+
+
+ """
+
+ def __init__(self, transformer, biometric_algorithm):
+ self.transformer = transformer
+ self.biometric_algorithm = biometric_algorithm
+
+ def __call__(
+ self,
+ background_model_samples,
+ biometric_reference_samples,
probe_samples,
- biometric_references,
- transformer,
- biometric_algorithm,
- allow_scoring_with_all_biometric_references,
- )
+ allow_scoring_with_all_biometric_references=False,
+ ):
+ logger.info(
+ f" >> Vanilla Biometrics: Training background model with pipeline {self.transformer}"
+ )
+ # Training background model (fit will return even if samples is ``None``,
+ # in which case we suppose the algorithm is not trainable in any way)
+ self.transformer = self.train_background_model(background_model_samples)
-def train_background_model(background_model_samples, transformer):
- # background_model_samples is a list of Samples
+ logger.info(
+ f" >> Creating biometric references with the biometric algorithm {self.biometric_algorithm}"
+ )
- # We might have algorithms that has no data for training
- if len(background_model_samples) <= 0:
- logger.warning(
- "There's no data to train background model."
- "For the rest of the execution it will be assumed that the pipeline is stateless."
+ # Create biometric samples
+ biometric_references = self.create_biometric_reference(
+ biometric_reference_samples
)
- return transformer
- transformer = transformer.fit(background_model_samples)
- return transformer
+ logger.info(
+ f" >> Computing scores with the biometric algorithm {self.biometric_algorithm}"
+ )
+ # Scores all probes
+ return self.compute_scores(
+ probe_samples,
+ biometric_references,
+ allow_scoring_with_all_biometric_references,
+ )
-def create_biometric_reference(
- biometric_reference_samples, transformer, biometric_algorithm
-):
- biometric_reference_features = transformer.transform(biometric_reference_samples)
+ def train_background_model(self, background_model_samples):
+ # background_model_samples is a list of Samples
- biometric_references = biometric_algorithm.enroll_samples(
- biometric_reference_features
- )
+ # We might have algorithms that has no data for training
+ if len(background_model_samples) <= 0:
+ logger.warning(
+ "There's no data to train background model."
+ "For the rest of the execution it will be assumed that the pipeline is stateless."
+ )
+ return self.transformer
- # models is a list of Samples
- return biometric_references
+ return self.transformer.fit(background_model_samples)
+ def create_biometric_reference(self, biometric_reference_samples):
+ biometric_reference_features = self.transformer.transform(
+ biometric_reference_samples
+ )
-def compute_scores(
- probe_samples,
- biometric_references,
- transformer,
- biometric_algorithm,
- allow_scoring_with_all_biometric_references=False,
-):
+ biometric_references = self.biometric_algorithm.enroll_samples(
+ biometric_reference_features
+ )
- # probes is a list of SampleSets
- probe_features = transformer.transform(probe_samples)
+ # models is a list of Samples
+ return biometric_references
- scores = biometric_algorithm.score_samples(
- probe_features,
+ def compute_scores(
+ self,
+ probe_samples,
biometric_references,
- allow_scoring_with_all_biometric_references=allow_scoring_with_all_biometric_references,
- )
+ allow_scoring_with_all_biometric_references=False,
+ ):
+
+ # probes is a list of SampleSets
+ probe_features = self.transformer.transform(probe_samples)
+
+ scores = self.biometric_algorithm.score_samples(
+ probe_features,
+ biometric_references,
+ allow_scoring_with_all_biometric_references=allow_scoring_with_all_biometric_references,
+ )
+
+ # scores is a list of Samples
+ return scores
+
+
+def dask_vanilla_biometrics(pipeline, npartitions=None):
+ """
+ Given a :py:class:`VanillaBiometrics`, wraps :py:meth:`VanillaBiometrics.transformer` and
+ :py:class:`VanillaBiometrics.biometric_algorithm` with Dask delayeds
+
+ Parameters
+ ----------
+
+ pipeline: :py:class:`VanillaBiometrics`
+ Vanilla Biometrics based pipeline to be dasked
+
+
+ npartitions: int
+ Number of partitions for the initial `Dask.bag`
+ """
+
+ from bob.pipelines.mixins import estimator_dask_it, mix_me_up
+ from bob.bio.base.pipelines.vanilla_biometrics.mixins import BioAlgDaskMixin
+
+ transformer = estimator_dask_it(pipeline.transformer, npartitions=npartitions)
+ biometric_algorithm = mix_me_up([BioAlgDaskMixin], pipeline.biometric_algorithm)
- # scores is a list of Samples
- return scores
+ return VanillaBiometrics(transformer, biometric_algorithm)
diff --git a/bob/bio/base/script/vanilla_biometrics.py b/bob/bio/base/script/vanilla_biometrics.py
index 14b910352c26ad1713b4e6329caa1c7ac62446e6..ea0ad30ec52ce3ea7f6937b890ba2c47f0f97f43 100644
--- a/bob/bio/base/script/vanilla_biometrics.py
+++ b/bob/bio/base/script/vanilla_biometrics.py
@@ -57,21 +57,13 @@ TODO: Work out this help
entry_point_group="bob.pipelines.config", cls=ConfigCommand, epilog=EPILOG,
)
@click.option(
- "--transformer",
- "-e",
+ "--pipeline",
+ "-p",
required=True,
cls=ResourceOption,
- entry_point_group="bob.pipelines.transformer",
+ entry_point_group="bob.pipelines.pipeline",
help="Feature extraction algorithm",
)
-@click.option(
- "--algorithm",
- "-a",
- required=True,
- cls=ResourceOption,
- entry_point_group="bob.bio.algorithm", # This should be linked to bob.bio.base
- help="Biometric Algorithm (class that implements the methods: `fit`, `enroll` and `score`)",
-)
@click.option(
"--database",
"-d",
@@ -106,7 +98,7 @@ TODO: Work out this help
)
@verbosity_option(cls=ResourceOption)
def vanilla_biometrics(
- transformer, algorithm, database, dask_client, groups, output, **kwargs
+ pipeline, database, dask_client, groups, output, **kwargs
):
"""Runs the simplest biometrics pipeline.
@@ -152,7 +144,7 @@ def vanilla_biometrics(
"""
- from bob.bio.base.pipelines.vanilla_biometrics.pipeline import biometric_pipeline
+ from bob.bio.base.pipelines.vanilla_biometrics.pipeline import VanillaBiometrics
import dask.bag
import itertools
import os
@@ -174,14 +166,11 @@ def vanilla_biometrics(
else False
)
- result = biometric_pipeline(
- database.background_model_samples(),
- biometric_references,
- database.probes(group=group),
- transformer,
- algorithm,
- allow_scoring_with_all_biometric_references=allow_scoring_with_all_biometric_references
- )
+ result = pipeline(database.background_model_samples(),
+ biometric_references,
+ database.probes(group=group),
+ allow_scoring_with_all_biometric_references=allow_scoring_with_all_biometric_references
+ )
if isinstance(result, dask.bag.core.Bag):
if dask_client is not None: