Update head and h-label format for hierarchical label classification

[sooahleex]

Summary

• Related to revamp model for h-cls

• Update head of h-label cls with CBAM head

• Update ground truth format of h-label cls

  • For example, this case has label like [-1, -1, -1, -1, -1, -1, -1, 2, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 7].
    This is example of cifar100, which has 100 subclasses and 20 superclasses.
    

• Only consider multi-class case, so for multi-label case, I will cover later.

How to test

• Update unit test and dataset for it

Checklist

• I have added unit tests to cover my changes.​
• I have added integration tests to cover my changes.​
• I have ran e2e tests and there is no issues.
• I have added the description of my changes into CHANGELOG in my target branch (e.g., CHANGELOG in develop).​
• I have updated the documentation in my target branch accordingly (e.g., documentation in develop).
• I have linked related issues.

License

• I submit my code changes under the same Apache License that covers the project.
  Feel free to contact the maintainers if that's a concern.
• I have updated the license header for each file (see an example below).

# Copyright (C) 2024 Intel Corporation
# SPDX-License-Identifier: Apache-2.0

[harimkang]

Looks good, I left a few comments, and could you please write some more unit-tests for the classes added to hlabel_cls_head.py?

Also, it would be great if you could summarize and share the results of experiments comparing the old head to the new head. (cifar100)

• @chuneuny-emily I think we should use cifar100 that sooah used in experiments here for H-label cls large performance tests, what do you think? (There is currently no Large-data case for H-label.)

[harimkang]

LGTM, I left a few minor comments.

[eunwoosh]

Thanks for your work. I left a minor comment. Please take a look.