This section is under development
The generated output from the deployed model can be consumed by custom business logic flows built by users, so that action can be taken upon the inference results in an automated way.
TODO: include boilerplate flows by using Node-RED for standard use cases like counting, measuring polygon/bounding box size and orientation, filtering labels or confidence thresholds