Implement YoloV11 OBB detection for RKNN JNI #2105
Conversation
github-actions
bot
added
documentation
This comment was marked as resolved.
This comment was marked as resolved.
|
Also, lint please 😄 |
| :items=" | ||
| useSettingsStore().general.supportedBackends?.includes('RKNN') | ||
| ? ['YOLOv5', 'YOLOv8', 'YOLO11'] | ||
| ? ['YOLOv5', 'YOLOv8', 'YOLO11', 'YOLO11OBB'] |
There was a problem hiding this comment.
Let's try and be consistent with how we use the v in yolov11. I'd prefer to do it the same across everywhere, and we currently have YOLOv11.
There was a problem hiding this comment.
Oh yeah I noticed that too!
For the other places I did YoloV11OBB, but for that specific place I did "YOLO11OBB" because the "YOLO11" behind it didn't have the "V"
Should we change both?
There was a problem hiding this comment.
For the other places using the v works fine, our goal is to keep it consistent. According to the branding standards, it should be yolo11 but that's what they get for changing it.
To whit, YOLOV11OBB please!
There was a problem hiding this comment.
I agree we should keep it consistent, but I was referring to:
['YOLOv5', 'YOLOv8', 'YOLO11', 'YOLO11OBB']
We should probably change Yolo11 to YoloV11 there as well
This comment was marked as resolved.
This comment was marked as resolved.
samfreund
force-pushed
the
yolo11obb
branch
2 times, most recently
from
d1c428e to
157a615
Compare
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
There was a problem hiding this comment.
I got it working! Just some cleanup stuff now.
-
I'd like to see only the bounding box for the obb displayed in the processed stream, I don't think we need the non-rotated one as well.
-
I saw you added the note about YOLO11-OBB being supported to the opi docs, I would love if you could write a little explanation of how to use this as well for the docs.
|
Also, if you're willing to do a writeup on the steps you took to convert the OBB model and how they differ from the typical detection model, that would be really helpful. |
|
Wait by the way, thank you for all your help in fixing some of the formatting and stuff up. I'm currently at university so I have to wait until I get back home before I can access the Orange PI. I noticed how it draws both the axis-aligned BB and OBB, but I didn't add the axis aligned one. It seems how Photonvision works is whenever its drawing a contour, it overlays the minimum rectangle over it as well. Do you have any ideas on how I could fix that ? |
In the future, you can do it yourself!
Drawing 2d targets is done here, I would guess that's where you would fix it. |
2 participants
Description
Implemented YoloV11 Oriented Bounding Box detection so recognized bounding boxes also include an angle theta describing their rotation. Helpful when the overall axis aligned BB is too inaccurate or big. Only supports RKNN JNI for now. (Corresponding PR: PhotonVision/rknn_jni#28)
Meta
Merge checklist: