i am testing roboflow workflow to run key-point detection and upload the results to a project, i.e. trying to create a annotator helper for my datasets.
the problem i am facing is that the images uploaded (through roboflow dataset upload ) does contain a bounding box but are missing the keypoints. the images appear in the project as “test-workflow-annotation”. i knowfrom the json and visualization that model correctly predicts the key-points.
the workflow contains the blocks: inputs - keypoint detection model - roboflow dataset upload - keypoint visualization - outputs
project: test-workflow-annotation
model: How to Use the digital-gauge Keypoint Detection API
workflow: yolov8n-pose digital gauge