Reliability

Table of contents

analyze_reliability()
analyze_reliability_for_categories()

analyze_reliability()

It provides the reliability analysis by showing the distribution of the proposals among different confidence values and the corresponding confidence calibration.

Parameters

bins: int, optional; The number of bins the confidence values are split into.
(default is 10)
show: bool, optional; Indicates whether the plot should be shown or not. If False, returns the results as dict.
(default is True)

Example

Classification

from odin.classes import AnalyzerClassification

my_analyzer = AnalyzerClassification("my_classifier_name", my_classification_dataset)
my_analyzer.analyze_reliability()

analyze_reliability_cl_output

Localization

from odin.classes import AnalyzerLocalization

my_analyzer = AnalyzerLocalization("my_detector_name", my_localization_dataset)
my_analyzer.analyze_reliability()

analyze_reliability_loc_output

Tasks supported

Binary Classification	Single-label Classification	Multi-label Classification	Object Detection	Instance Segmentation
yes	yes	yes	yes	yes

analyze_reliability_for_categories()

It provides the reliability analysis by showing the distribution of the proposals among different confidence values and the corresponding confidence calibration for each category.

Parameters

categories: list, optional; If not specified, it performs the analysis on the entire data set, otherwise it performs a per-category analysis.
(default is None)
bins: int, optional; The number of bins the confidence values are split into.
(default is 10)
show: bool, optional; Indicates whether the plot should be shown or not. If False, returns the results as dict.
(default is True)

Example

Classification

from odin.classes import AnalyzerClassification

my_analyzer = AnalyzerClassification("my_classifier_name", my_classification_dataset)
my_analyzer.analyze_reliability_for_categories()

analyze_reliability_cat_cl_output

N.B. As example, it is shown only the output of a single category, but the analysis is performed for all the categories selected.

Localization

from odin.classes import AnalyzerLocalization

my_analyzer = AnalyzerLocalization("my_detector_name", my_localization_dataset)
my_analyzer.analyze_reliability_for_categories()

analyze_reliability_cat_loc_output

N.B. As example, it is shown only the output of a single category, but the analysis is performed for all the categories selected.

Tasks supported

Binary Classification	Single-label Classification	Multi-label Classification	Object Detection	Instance Segmentation
no	yes	yes	yes	yes