add Extended-IOU loss implementation

[Wwupup]

Motivation

In paper: A Systematic IoU-Related Method: Beyond Simplified Regression for Better Localization of TIP2021, we propose a new Extended-IOU loss, which outperforms the commonly used GIOU and DIOU in object detection.

Modification

add implementation of the proposed Extended-IOU Loss (EIOU).

[Wwupup]

Hello, this is the first time I submit code to MMDetection. I've checked the failure details but can't figure out what's wrong, please help me check if it's convenient.

[jbwang1997]

Hello @Wwupup. Thanks for your contributions. You can first follow the CONTRIBUTING.md to install pre-commit and fix the lint error. Besides, it's better to also provide a config file in this PR.

[hhaAndroid]

@Wwupup Can you also create a PR to the dev-3.x branch?

[Wwupup]

Hello @Wwupup. Thanks for your contributions. You can first follow the CONTRIBUTING.md to install pre-commit and fix the lint error. Besides, it's better to also provide a config file in this PR.

Hi, I have followed the manual guidelines before submitting my PR. But I did pre-commit locally and it shows all passed. But, error in Lint

[RangiLyu]

We updated the precommit version recently. You can rebase your branch to the latest dev branch to solve this problem.

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Comparison is base (9d3e162) 64.12% compared to head (e3b5caa) 64.22%. Report is 77 commits behind head on dev.

Additional details and impacted files

@@            Coverage Diff             @@
##              dev    #9076      +/-   ##
==========================================
+ Coverage   64.12%   64.22%   +0.10%     
==========================================
  Files         361      361              
  Lines       29536    29585      +49     
  Branches     5022     5027       +5     
==========================================
+ Hits        18939    19002      +63     
+ Misses       9580     9579       -1     
+ Partials     1017     1004      -13     

Flag	Coverage Δ
unittests	64.20% <100.00%> (+0.08%)	:arrow_up:

Flags with carried forward coverage won't be shown. Click here to find out more.

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.