[Python] How to modify YOLOv9 to handle additional parameters in ground truth?

I am working on configuring YOLOv9 to handle a custom ground truth format. The ground truth file I am using has additional parameters (such as landmarks) that I need to include in the training process, but I'm facing issues in integrating these extra columns into the existing YOLOv9 configuration. What i need to change in Yolo?

I expect that my new Yolov9 model could train with new ground truth file

Continue reading...