GuidesDatasets and Models

Object Detection Inputs

Prepare YOLO or COCO bounding-box data and a compatible detection Model.

Object detection compares predicted axis-aligned boxes and classes with the Dataset annotations. Knotest supports YOLO Object Detection and COCO Object Detection Dataset formats.

Preparation checklist

  • Confirm that every annotation resolves to a readable image.
  • Validate box coordinates and reject zero-area or out-of-bounds geometry.
  • Confirm Dataset split selection before import.
  • Preserve the Dataset class ID to label mapping.
  • Configure the Model with the identical class order.
  • Record confidence threshold, NMS behavior, maximum detections, and input size.

For .pt YOLO Models, choose the YOLO backend and Object Detection task during Model import. Do not silently change inference settings between Prediction Sets; they are part of provenance.

Helmet YOLO26 detector Model settings with the helmet class and 640 by 640 input dimensions.

Watch: Review model settingsOpen the imported detector and inspect its class order and input dimensions.Silent demo · Subtitles follow page language · Full screen for detail

Classification Inputs

Prepare and import an image classification Dataset and compatible Model.

Segmentation Inputs

Prepare YOLO Segmentation or COCO Instance Segmentation inputs.

On this page

Preparation checklist