Skip to content

fix(bev_fusion): revert the addition of TRAFFIC_CONE and BARRIER from the config to match the current released model version - #13506

Merged
xmfcx merged 1 commit into
autowarefoundation:mainfrom
a-maumau:mau/perception/bevfusion/revert_configs
Oct 18, 2026
Merged

xmfcx merged 1 commit into
autowarefoundation:mainfrom
a-maumau:mau/perception/bevfusion/revert_configs

Conversation

@a-maumau

Copy link
Copy Markdown
Contributor

Description

This PR will revert the changes that was made in #12732 to match the current published model version.

The configs in the autoware launch should be synced after this PR merged.

Related links

Parent Issue:

How was this PR tested?

Tested with sample rosbag and confirmed the output from bevfusion has no issues.

Notes for reviewers

None.

Interface changes

None.

Effects on system behavior

None.

…tch the current released model version

Signed-off-by: Masaki Baba <masaki.baba.2@tier4.jp>
@github-actions

Copy link
Copy Markdown

Thank you for contributing to the Autoware project!

🚧 If your pull request is in progress, switch it to draft mode.

Please ensure:

@github-actions github-actions Bot added the component:perception Advanced sensor data processing and environment understanding. (auto-assigned) label Oct 16, 2026
@xmfcx xmfcx added the run:build-and-test-differential Mark to enable build-and-test-differential workflow. (used-by-ci) label Oct 16, 2026
@xmfcx
xmfcx self-requested a review October 16, 2026 15:06

@xmfcx xmfcx left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you @a-maumau -san. I compared the files with the released t4base_120m/v2 model. After this PR, ml_package_bevfusion_lidar.param.yaml, ml_package_bevfusion_camera_lidar.param.yaml, and detection_class_remapper.param.yaml are byte-identical to the v2.0 files on Hugging Face. This fixes #13023.

After the merge, please also sync bevfusion_lidar.param.yaml and bevfusion_camera_lidar.param.yaml in autoware_launch, as you wrote. They still have the 7-class yaw_norm_thresholds and the TRAFFIC_CONE and BARRIER scores.

@xmfcx
xmfcx merged commit d16d0f2 into autowarefoundation:main Oct 18, 2026
41 of 42 checks passed
@github-project-automation github-project-automation Bot moved this from To Triage to Done in Software Working Group Oct 18, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

component:perception Advanced sensor data processing and environment understanding. (auto-assigned) run:build-and-test-differential Mark to enable build-and-test-differential workflow. (used-by-ci)

Projects

Archived in project

Development

Successfully merging this pull request may close these issues.

2 participants