metrics[7]: speedup binclf with numba [GSoC23 cont'd]

[jpcbertoldo]

Description

This is a continuation of Anomaly Segmentation Metrics for Anomalib — GSoC 2023 @ OpenVINO.

Updated PRs at https://gist.github.com/jpcbertoldo/12553b7eaa97cfbf3e55bfd7d1cafe88?permalink_comment_id=4681988#gistcomment-4681988

---

Accelerate the computation of binary classification curves per image by using numba.

The functions

_binclf_curve_numba (for one image)
_binclf_curves_numba_parallel (for many images)

are respectively equivalent to

• __binclf_curves_ndarray_itertools (for one image)
• _binclf_curves_ndarray_itertools (for many images)

ATTENTION: new requirement numba!

Changes

Describe the changes you made

• Bug fix (non-breaking change which fixes an issue)
• Refactor (non-breaking change which refactors the code base)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update

Checklist

Ensure that you followed the following

• I have added a summary of my changes to the CHANGELOG (not for minor changes, docs and tests).
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas (If applicable)
• I have made corresponding changes to the documentation (If applicable)
• I have added tests that prove my fix is effective or that my feature works (If applicable)

[jpcbertoldo]

Replaced by openvinotoolkit#1557 to catch up with v1.