feat(autoware_ml): add t4segmetric#195
Merged
amadeuszsz merged 25 commits intotier4:mainfrom Mar 25, 2026
Merged
Conversation
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Contributor
There was a problem hiding this comment.
Pull request overview
Adds a shared 3D segmentation evaluation core (t4_seg_eval) and integrates it into both PTv3 evaluation/test flows and FRNet/MMEngine via a new T4SegMetric, including optional BEV distance-range breakdown and confusion-matrix visualization/logging.
Changes:
- Introduced
autoware_ml.segmentation3d.evaluation.functional.t4_seg_evalwith unified metrics (IoU/acc/precision/recall/F1), confusion matrices, and optional range bucketing. - Added
autoware_ml.segmentation3d.evaluation.metrics.T4SegMetric(MMEngineBaseMetric) wrapping the functional core and logging CM images. - Updated PTv3 evaluator/tester and FRNet configs to use the shared evaluator + distance-range options.
Reviewed changes
Copilot reviewed 11 out of 11 changed files in this pull request and generated 9 comments.
Show a summary per file
| File | Description |
|---|---|
| projects/PTv3/engines/test.py | Switches PTv3 test-time evaluation to t4_seg_eval and adds TensorBoard + CM/range logging. |
| projects/PTv3/engines/hooks/evaluator.py | Refactors PTv3 validation evaluator to t4_seg_eval + range/CM logging. |
| projects/PTv3/configs/semseg-pt-v3m1-0-t4dataset.py | Adds metric_options.distance_ranges for PTv3 runs. |
| projects/FRNet/configs/t4dataset/frnet_1xb8_t4dataset-qt128-seg.py | Uses T4SegMetric with distance ranges and updates hooks config. |
| projects/FRNet/configs/t4dataset/frnet_1xb8_t4dataset-ot128-seg.py | Uses T4SegMetric with distance ranges and updates hooks config. |
| projects/FRNet/configs/nuscenes/frnet_1xb4_nus-seg.py | Updates hooks config to include LoggerHook and save_best='miou'. |
| autoware_ml/segmentation3d/evaluation/functional/t4_seg_eval.py | New unified functional evaluator + plotting/helpers. |
| autoware_ml/segmentation3d/evaluation/metrics/t4_seg_metric.py | New MMEngine metric adapter with optional CM image logging + submission export. |
| autoware_ml/segmentation3d/evaluation/init.py | Exposes evaluation public API (t4_seg_eval, T4SegMetric, helpers). |
| autoware_ml/segmentation3d/evaluation/functional/init.py | Exposes functional helpers. |
| autoware_ml/segmentation3d/evaluation/metrics/init.py | Exposes T4SegMetric. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
You can also share your feedback on Copilot code review. Take the survey.
autoware_ml/segmentation3d/evaluation/functional/t4_seg_eval.py
Outdated
Show resolved
Hide resolved
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp> Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
…mplicit spconv load Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
…mplicit spconv load Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Collaborator
Author
|
@codex review |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: e45c323055
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
…SegMetric Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
…etric Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
Signed-off-by: Amadeusz Szymko <amadeusz.szymko.2@tier4.jp>
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.
Summary
Add
T4SegMetric- a unified segmentation evaluation module shared by PTv3 and FRNet. It computes per-class and mean IoU, precision, recall, F1, and optionally BEV distance-range-based metrics with confusion matrix visualization and TensorBoard logging.Change point
t4_seg_evalfunctional core (autoware_ml/segmentation3d/evaluation/functional/t4_seg_eval.py): shared evaluation logic producingSegEvalResult(metrics dict, confusion matrix, per-range CMs). Includesfast_hist,per_class_iou/precision/recall/f1,plot_confusion_matrix, and range-based BEV bucketing.T4SegMetricMMEngine metric (autoware_ml/segmentation3d/evaluation/metrics/t4_seg_metric.py):BaseMetricsubclass wrappingt4_seg_evalfor FRNet runner integration with optional confusion matrix image logging.projects/PTv3/engines/hooks/evaluator.py,projects/PTv3/engines/test.py): replaced inline metric computation witht4_seg_eval, added range-based metric and confusion matrix TensorBoard logging.projects/FRNet/configs/t4dataset/frnet_1xb8_t4dataset-{ot,qt}128-seg.py): switched fromSegMetrictoT4SegMetricwithdistance_ranges.Note
Test performed