Repository navigation
fix(clean): preserve exact tracked source frame grids - #136
Closed
austinjiann wants to merge 1 commit into
Closed
austinjiann wants to merge 1 commit into
austinjiann wants to merge 1 commit into
Conversation
This was referenced Nov 21, 2026
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A source-aligned tracking run can retain an endpoint that the FPS filter omits. The city clip had 94 tracked frames, including source frame185, while a fresh12fps filter produced93. Cleaning stopped locally despite valid masks and exact timestamps.
Explicit-source mask archives now validate their count against the decoded recorded selection, preserving every source index and PTS. Legacy archives still require the FPS-filter count. Source hashes, dimensions, timestamps and complete declared track-mask inventory remain mandatory; malformed explicit metadata cannot fall back to legacy sampling.
Validation: 2,756 Python tests and
bun run format:checkpassed. Fractional-FPS regression exercises94vs93, endpoint preservation and refusal before provider submission for corrupt metadata/missing masks. The real city replay produced94 masks with no missing samples; beginning/middle/end overlays were inspected. No visual-quality thresholds changed. Background pedestrians and some bag edges remain outside these masks.Stacked on #135. Actual paid cleaning is being exercised separately; local mask validation alone is not a clean-scene quality verdict.