Creative Technology · Computer Vision

AI-Powered Object Detection

Upload one image. A pretrained AI model finds what is in it and where each object is.

Runs directly in your browser · no GPU · no server · no training

Choose an image

Confidence threshold 50%

Choose an image to begin.

01 · IMAGE

What does the AI see?

0 objects

Your image will appear here.

02 · RESULT

Detected objects

Run the detector to see object counts.

Object	Confidence
No results yet.

03 · WHAT JUST HAPPENED?

Image→Pretrained detector→Boxes + labels + confidence

This demo uses Xenova/detr-resnet-50 through Transformers.js. The model is downloaded to the browser on the first run and inference happens on the user's device.