I’m trying to build a web app that uses a keypoint detection model. For that, I’m using the roboflow.js directly via cdn.
<script src="https://cdn.roboflow.com/0.2.26/roboflow.js"></script>
When running model(...).detect(image)
with a simple detection model (rock paper scissors for example) it works, but when running for a keypoint detection model it returns some weird predictions that dont make any sense.