Enable configurable confidence threshold for ov export and inference

[jaegukhyun]

Summary

This PR includes enabling to control confidence threshold during eval(torch, ov models) and export.
From now on user can control confidence threshold via turning off result_based_threshold and setting confidence threshold through cli or configuration.yaml.

This is for #2291, a request for changing variables during evaluation and export procedure. Other variables such as input sizes and iou_threshold will be controlled after those variables can be controlled by configurable parameters.

How to test

0. Train object detection model
1. Export model through $ otx export -o outputs/exported/origin/
2. Eval model throuhg $ otx eval --load-weights outputs/exported/origin/openvino.xml
3. Eval model through $ otx eval --load-weights outputs/exported/origin/openvino.xml params --postprocessing.result_based_confidence_threshold false --postprocessing.confidence_threshold 0.95
4. Export model through $ otx export -o outputs/exported/0.95/ params --postprocessing.result_based_confidence false --postprocessing.confidence_threshold 0.95
5. Eval model through $ otx eval --load-weights outputs/exported/0.95/openvino.xml

You can see largely degraded performance in step 3 and results of step 3 and step 5 are same.

Checklist

• I have added unit tests to cover my changes.​
• I have added integration tests to cover my changes.​
• I have added e2e tests for validation.
• 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) 2023 Intel Corporation
# SPDX-License-Identifier: Apache-2.0

[eunwoosh]

LGTM but there is a minor typo. Could you fix it?

[sungmanc]

You can see largely degraded performance in step 4 and results of step 4 and step 6 are same.

This sentence is confusing. I think step 2 > (step 3 = step 5) is what you intended, is it right?

[goodsong81]

Generally looks good, but could you review the review comments including mine?

[jaegukhyun]

You can see largely degraded performance in step 4 and results of step 4 and step 6 are same.

This sentence is confusing. I think step 2 > (step 3 = step 5) is what you intended, is it right?

Yeap that' right, I changed the comments.