This repo uses YOLOv8-pose to train the 14th object of the datastes of T-less,then predicts its key-points and estimates its pose with a deep camera.
Step1. Label the 14th object with keypoints by Labelme. Step2. Transform the keypoints' informations to training of YOLOv8-pose. Step3. Predict the key-points of 14th objects. Step4. Estimate the object's pose by the keypoints with deep camera.