perf(type_adaptation): Cuda type adaptation

[Owen-Liuyuxuan]

PR Type

• New Feature

Related Links

CUDA Adaptation

https://github.com/Owen-Liuyuxuan/autoware_type_adapters

YOLOX Adapatation

tier4/autoware_universe#1637

Description

Using Type Adaptation to allow transferring messages while keeping them in GPU.

Build the PR branch along with autoware_type_adapters in an anvil Orin, we can get things running.

• Implement type adapters for NPP rectification and Jetson compression.
• I have tested that it correctly rectified and compressed images on an Anvil with four 20Hz 1920*1080 images.
• We can use the node interface as if it is still a normal image.
• But as we also implement YoloX interface/process with type_adapters, we could obtain much more significantly improvements.

Review Procedure

Remarks

Pre-Review Checklist for the PR Author

PR Author should check the checkboxes below when creating the PR.

• Assign PR to reviewer

Checklist for the PR Reviewer

Reviewers should check the checkboxes below before approval.

• Commits are properly organized and messages are according to the guideline
• (Optional) Unit tests have been written for new behavior
• PR title describes the changes

Post-Review Checklist for the PR Author

PR Author should check the checkboxes below before merging.

• All open points are addressed and tracked via issues or tickets

CI Checks

• Build and test for PR: Required to pass before the merge.

[Owen-Liuyuxuan]

This is a common bug fix, not directly related to type adaptation, but needed.