Enhance Evaluation Metrics Logging and Add PR Curves to TensorBoard #25
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
- Introduced functions to create and log precision-recall curves to TensorBoard. - Added methods to extract precision-recall data from COCO evaluation results. - Implemented F1 score calculation based on precision and recall values. - Updated the evaluate function to include logging of precision-recall curves and F1 scores for both bounding box and segmentation metrics. - Enhanced imports to include necessary libraries for plotting and numerical operations.
- Removed redundant comments in the Trainer class for clarity. - Improved logging structure in the Trainer class for better readability. - Updated precision-recall curve logging paths in the det_engine to follow a consistent naming convention. - Enhanced evaluation metrics logging to include top-level metrics for mAP and mAR, along with F1 score calculations.
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Outdated suggestions cannot be applied.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 significantly enhances the evaluation metrics logged to TensorBoard, providing more detailed insights into model performance, including Precision-Recall (PR) curves.
Key Changes:
src/deimkit/engine/solver/det_engine.py):metrics-AP/,metrics-AR/). Includes breakdowns by IoU threshold and object area size.metrics-F1/.top-level-metrics/category in TensorBoard for quick access to primary mAP (bbox_stats[0]), mAR (bbox_stats[8]), and the corresponding F1 score.src/deimkit/engine/solver/det_engine.py):metrics-PR/category.matplotlibas a dependency for plotting.Screenshot
