[Feature] Add two traditional filters to smooth poses

[ailingzengzzz]

Motivation

Smooth poses.

Modification

1. Add Smoother module for pose smoothing, which applies a temporal filter on pose sequence.
2. Support 3 filters: Savitzky-Golay 2d filter, Gaussian 1d filter, and One-Euro filter (derived from the existing implementation in MMpose)
3. Update inference APIs, demos, and Webcam APIs.
4. Add unit tests of smoother and filters
5. Fix the bug reported in Mistake in one_euro_filter.py #1176

Use Cases

Init smoother

from mmpose.core import Smoother

smoother = Smoother(filter_cfg='configs/_base_/filters/one_euro.py', keypoint_dim=2)

Online smoothing (frame-by-frame)

for frame in video:
    # pose estimation
    results = inference_top_down_pose_model(model, frame, ...)
    # pose tracking
    results, next_id = get_track_id(results, last_results, next_id, ...)
    # smoothing
    results_smoothed = smoother.smooth(results)

Offline smoothing

results_sequence = []

for frame in video:
    # pose estimation
    results = inference_top_down_pose_model(model, frame, ...)
    # pose tracking
    results, next_id = get_track_id(results, last_results, next_id, ...)

    results_sequence.append(results)

results_sequence_smoothed = smoother.smooth(results_sequence)

BC-Breaking

The arguments use_one_euro and fps in function get_track_id() will be deprecated in the future. Deprecation warnings have been added in this PR.

Checklist

Before PR:

• I have read and followed the workflow indicated in the CONTRIBUTING.md to create this PR.
• Pre-commit or linting tools indicated in CONTRIBUTING.md are used to fix the potential lint issues.
• Bug fixes are covered by unit tests, the case that causes the bug should be added in the unit tests.
• New functionalities are covered by complete unit tests. If not, please add more unit tests to ensure correctness.
• The documentation has been modified accordingly, including docstring or example tutorials.

After PR:

• CLA has been signed and all committers have signed the CLA in this PR.

[CLAassistant]

All committers have signed the CLA.

[jin-s13]

Pls follow https://github.com/open-mmlab/mmpose/blob/master/.github/CONTRIBUTING.md to setup pre-commit.

[codecov]

Codecov Report

Merging #1127 (b77a93c) into dev-0.25 (c09a693) will increase coverage by 0.03%.
The diff coverage is 96.41%.

❗ Current head b77a93c differs from pull request most recent head 605bbae. Consider uploading reports for the commit 605bbae to get more accurate results

@@             Coverage Diff              @@
##           dev-0.25    #1127      +/-   ##
============================================
+ Coverage     83.44%   83.47%   +0.03%     
============================================
  Files           205      212       +7     
  Lines         16625    16821     +196     
  Branches       2976     3010      +34     
============================================
+ Hits          13872    14042     +170     
- Misses         2001     2012      +11     
- Partials        752      767      +15     

Flag	Coverage Δ
unittests	83.46% <96.41%> (+0.09%)	⬆️

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

Impacted Files	Coverage Δ
mmpose/apis/inference_tracking.py	59.05% <50.00%> (-0.30%)	⬇️
mmpose/core/post_processing/smoother.py	94.59% <94.59%> (ø)
...ost_processing/temporal_filters/one_euro_filter.py	98.07% <98.07%> (ø)
mmpose/core/post_processing/__init__.py	100.00% <100.00%> (ø)
mmpose/core/post_processing/one_euro_filter.py	100.00% <100.00%> (ø)
.../core/post_processing/temporal_filters/__init__.py	100.00% <100.00%> (ø)
...e/core/post_processing/temporal_filters/builder.py	100.00% <100.00%> (ø)
...se/core/post_processing/temporal_filters/filter.py	100.00% <100.00%> (ø)
...ost_processing/temporal_filters/gaussian_filter.py	100.00% <100.00%> (ø)
...rocessing/temporal_filters/savizky_golay_filter.py	100.00% <100.00%> (ø)
... and 4 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c09a693...605bbae. Read the comment docs.

[jin-s13]

Do we need any description here?

[ly015]

This is part of the newly reimplemented One-Euro filter. The one that will be deprecated is in mmpose/core/post_processing/one_euro_filter.py.

[OpenMMLab-Assistant-004]

Hi @ailingzengzzz ！First of all, we want to express our gratitude for your significant PR in the mmpose project. Your contribution is highly appreciated, and we are grateful for your efforts in helping improve this open-source project during your personal time. We believe that many developers will benefit from your PR.

We would also like to invite you to join our Special Interest Group (SIG) private channel on Discord, where you can share your experiences, ideas, and build connections with like-minded peers. To join the SIG channel, simply message moderator— OpenMMLab on Discord or briefly share your open-source contributions in the #introductions channel and we will assist you. Look forward to seeing you there! Join us ：https://discord.gg/raweFPmdzG

If you have WeChat，welcome to join our community on WeChat. You can add our assistant ：openmmlabwx. Please add "mmsig + Github ID" as a remark when adding friends：）
Thank you again for your contribution❤