mortal compass
Hi all - working on a program that needs to track keypoints using Yolo V8 Pose. I have it running under ONNX and getting fed the correct input data with OpenCV, but I'm not sure how to interpret it at all. I've been messing with the same 20-30 lines of code for hours and combing through google search results for example code.
Yolov8 pose outputs a bounding box, confidence, and 17 X,Y keypoints that show where human limbs are. I am getting an output tensor that is 1x56x8400. Where do I start with this data? Any way I try to parse it just looks like garbage 😦
