search

Qengineering / YoloV8-ncnn-Raspberry-Pi-4

YoloV8 for a bare Raspberry Pi 4 or 5
https://qengineering.eu/deep-learning-examples-on-raspberry-32-64-os.html
BSD 3-Clause "New" or "Revised" License
99 stars 7 forks source link

Yolov8 OBB #11

Closed kevinzezel closed 3 months ago

kevinzezel commented 3 months ago

Hi,

Do you have an example of how to run post-processing for a yolov8 OBB model?

Regards, Kevin

Qengineering commented 3 months ago

What are you referring to with post-processing? Usually, post-processing is preparing the outcome to be human understandable. In other words, make nice images with bounding boxes.

kevinzezel commented 3 months ago

You are right. I was referring to the part after inference in NCNN. Since YOLOv8 OBB works with angles in the boxes (X Y W H Theta), some functions like generate_proposals, nms_sorted_bboxes, etc., are different.

Thank you!

Qengineering commented 3 months ago

I see. Unfortunately, the current ncnn YoloV8 implementation doesn't support OBB. As far as I know, there is no YoloV8-OBB for ncnn available (yet).

kevinzezel commented 3 months ago

Thank you!

kevinzezel commented 3 months ago

I found one example:

https://github.com/FeiGeChuanShu/ncnn-android-yolov8/tree/main/ncnn-yolov8s-obb

Regards, Kevin