Detect Objects
Detect Objects
Section titled “Detect Objects”Find objects and their bounding boxes in an image, running locally in your browser (COCO-SSD on TensorFlow.js, DETR on transformers.js). Private, offline, no cost.
What it does
Section titled “What it does”Runs the image through the selected on-device detector and returns every object found — its label, confidence, and box — sorted by confidence and filtered by a threshold.
When to use it
Section titled “When to use it”Count or locate things in an image: how many people are in a photo, where a product sits, whether a frame contains a given object — then branch or annotate downstream.
Inputs and settings
Section titled “Inputs and settings”| Setting | Notes |
|---|---|
| Model | Any installed object-detection model, across engines. |
| Image | Image URL, data URL, or $binary from a previous node. |
| Confidence threshold | Drop detections below this score (default 0.5). |
| Max objects | Keep at most this many objects, highest score first (default 20). |
Outputs
Section titled “Outputs”Returns { objects, count, top }:
objects— array of{ label, score, box }, highest confidence first.boxholds the bounding-box coordinates.count— number of objects returned (after the threshold and Max objects).top— label of the most confident detection (ornull).
Troubleshooting
Section titled “Troubleshooting”- Nothing detected — lower the threshold, or use a model trained on the objects you expect.
- Fewer objects than expected in a busy image — raise Max objects.
- Empty picker — install an object-detection model first.