Running the pose model with empty frame when low confidence detector output. #138
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Collaborator
|
i can't really evaluate this, since i don't know how the pytorch model works, but it looks reasonable - doesn't seem like it should be possible to fall back to an array with a totally different shape dependent on quality of estimation |
MMathisLab
added a commit
that referenced
this pull request
Jan 13, 2026
* Update benchmark.py - changed benchmarking link; needs testing * fix #112 - don't subset cpuinfo (#114) * Introduce CI/CD for TensorFLow version (#122) * Update .gitignore * CI/CD update python version * CI/CD update actions versions * CI/CD update trigger events * CI/CD update MacOS version * dlclibrary set version to >=0.0.6 * Poetry lock * Poetry lock * Pyproject.toml update tensorflow installation * Poetry lock * Install specific tensorflow-io-gcs-filesystem for windows * Poetry lock * Update deprecated section name * Poetry lock * CI/CD test on python 3.11, 3.12, 3.13 as well * Update testing.yml - rename to main - add tables installation --------- Co-authored-by: Mackenzie Mathis <mathis@rowland.harvard.edu> * master --> main * Update testing.yml * Update benchmark.py * Update README & benchmarking script (#126) * Update README.md fix test instructions * Update check_install.py * Update check_install.py * Update check_install.py * Update pyproject.toml (#125) * Update pyproject.toml - adding support for python 3.11, 3.12, 3.13 * Update testing.yml * Update testing.yml * Update pyproject.toml * Update testing.yml * Update pyproject.toml * Update poetry.lock * Update poetry.lock * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update testing.yml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update testing.yml * Update pyproject.toml * Update pyproject.toml * Update poetry.lock * Update testing.yml * Update testing.yml * Update testing.yml - remove poetry (for now) * Update testing.yml * Update testing.yml * Update testing.yml * Update testing.yml * Update testing.yml * updated poetry.lock * updated * Update testing.yml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update testing.yml testing 3.11 * Update testing.yml * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * Update testing.yml * updated test for win32 * Update testing.yml * Update testing.yml * Update testing.yml * Update testing.yml * Fix download_benchmarking_data() * Ignore __MACOSX directories * Run tests on all PRs * tensorflow-io-gcs-filesystem - python=3.9 * Pin python to 3.10 version * WIP refactor benchmarking: __init__.py * WIP refactor benchmarking: remove dupplicates * WIP refactor benchmarking: warning * WIP refactor benchmarking: close live * WIP refactor benchmarking: resize and pixels * WIP refactor benchmarking: get_sys_info * Update benchmark.py (#123) * WIP refactor benchmarking: Delete dupplicate scripts * WIP refactor benchmarking: Introduce Engine * WIP refactor benchmarking: benchmark() method * WIP refactor benchmarking: benchmark_videos() * CI/CD: add torch installation * WIP refactor benchmarking: extract Engine * Refactor benchmarking: model_type in benchmark_videos() * Delete poetry lock * Fix frame_batch initialization in runner.py when low detection confidence (#138) --------- Co-authored-by: Mackenzie Mathis <mathis@rowland.harvard.edu> Co-authored-by: Jonny Saunders <sneakers-the-rat@protonmail.com> Co-authored-by: Artur <35294812+arturoptophys@users.noreply.github.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Addresses issue #137 with a quick-and-dirty trick:
when low-confidence prediction from detector -> pass a tensor of zeros to the pose model instead of raw frame.
This way this frame will not have valid predictions, but the DLCLive will not crash and continue with next frame.