Flesh out mediapipe model info #15
Merged
Conversation
This file contains 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
Update mediapipe_model_info.py
Update YOLO model info
…p process folder of videos
Co-authored-by: philipqueen <[email protected]>
…b.com/freemocap/skellytracker into philip/flesh_out_mediapipe_model_info
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Outdated suggestions cannot be applied.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.
This PR ensures that the model_info class for the mediapipe tracker contains all of the necessary information to replace the mediapipe model information held in Freemocap and its subrepos.
The goal is to have skellytracker manage all of the necessary model info for each image tracker. This allows freemocap and the downstream subrepos to be run independently of knowing any details about the tracker it's using. It also makes it easier to add new tracking models, as everything that needs to be added can be added in one place rather than having to updates multiple repos.
The only downside I see to the current approach is that we are doing a mix of inheriting the names from mediapipe, like in
hand_landmark_names = [landmark.name.lower() for landmark in mp_holistic.HandLandmark], and referencing the names with hard coded values:joint_hierarchy = {"hips_center": {"children": ["right_hip", "left_hip", "trunk_center"]}, ...}. This makes the code vulnerable to upstream changes coming from mediapipe, but is how the existing code works so I'm replicating it here.