Choose a source above to start
model not loaded
Live counts
COCO-SSD, 90 object classes, updating every frame.
No detections yet.
0
Total objects detected
Privacy by construction
TensorFlow.js loads the COCO-SSD model into your browser and runs every inference locally on your GPU/CPU
via WebGL. No frame, image, or detection result is ever sent to a server — there isn't one. The Python
reference implementation in python/counter.py runs the equivalent task offline with YOLOv8n
for a single image, if you'd rather run it as a batch script.