Skip to content
Work data is processed in your browser, not stored on our servers.
tools
AI Image Object Detector

AI Image Object Detector

Analyze a photo with DETR and display English labels, bounding boxes, scores, and counts for COCO's 80 object categories.

Focus mode
Image
Detect objects in a photo and draw bounding boxes around them. Processing stays in your browser.
90%
Detection results
80 COCO object classes (English labels).
Upload an image and select “Detect objects.”

Key features

  • Detects objects from the 80 COCO dataset categories
  • Overlays each detected position with a label, score, and bounding box
  • Summarizes total detections and counts by category
  • Adjusts the confidence threshold from 50% to 95%

How to use

  1. Choose or paste the image you want to inspect.
  2. Adjust the confidence threshold if needed and run detection.
  3. Review the boxes and per-category counts.

Examples

  • Survey a photo's contents

    Inspect the COCO categories and bounding boxes the model recognizes in an image.

  • Compare confidence thresholds

    Adjust the minimum score and compare which candidates and scores remain visible.

Privacy

The image is processed inside your browser for DETR inference. After you consent on first use, model files may be downloaded from Hugging Face and cached in browser storage.

Requirements

The first run requires consent to download about 43 MB of model files. Detection is limited to COCO categories and varies with the image and confidence threshold.

Common questions

What changes when I adjust the confidence threshold?
It changes the minimum score required to display a candidate. Lower values can show more boxes, including more false detections.
Can it find every object in a photo?
No. It detects candidates only from the model's COCO categories and can miss small, obscured, or unfamiliar objects.

Related tools