Apply activity-specific cutoff frequency to 2D keypoint filtering - #10
Open
Seeeeeyo wants to merge 1 commit into
Open
Apply activity-specific cutoff frequency to 2D keypoint filtering#10Seeeeeyo wants to merge 1 commit into
Seeeeeyo wants to merge 1 commit into
Conversation
Previously, the 2D keypoints used in the reprojection loss were Butterworth-filtered with the default 6 Hz cutoff before the activity (walking/squat/STS/other) was classified, so the activity-specific cutoff frequencies in params/parameters.yaml never actually reached the keypoint filter. Defer the filtering step until after activity detection so the intended per-activity cutoff is used. Co-authored-by: Cursor <cursoragent@cursor.com>
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
optimization.py) was previously always run with the function's default 6 Hz cutoff, because it ran before the activity (walking/treadmill/squat/STS/other) was classified.params/parameters.yaml(6 Hz walking/treadmill, 4 Hz squat/STS, 8 Hz other) were computed but only ever fed into an FFT-based smoothness loss term (loss_frequency) whose weight is absent from every weight preset, so they had no effect on the actual keypoint filtering (or on anything else, since that loss term is effectively dead given its zero/missing weight).Test plan
optimization.pyparses correctly (ast.parse) with the project'sopencapconda environment.run_optimizationend-to-end on a walking, squat/STS, and "other" sample video to confirm activity-specific filtering is applied and results remain reasonable.Made with Cursor