Skip to main content

Classification

This rule checks that the image classifies as specific labels.

Parameter table​

paramunitrequireddescription
confidencenoThe minimum confidence level required for the classification labels detection.
requiredLabelsyesAn array of classification labels to check for.

How to setup​

This rule evaluates whether an image contains classification labels. Key parameters include:

  • confidence: The minimum confidence level required for the classification labels detection.
  • requiredLabels: An array of classification labels to check for.

If no confidence level is specified, the default value is 30%. The rule will fail if the image does not contain all required labels.

Examples​

valuevalidconfidencerequiredLabelsdescription
Car✅80(Car, Person)Image should contain the labels 'Car' and 'Person' with a confidence level of 80% or higher.
Axe✅(Axe)Image should contain the label 'Axe'
Axe❌90(Avocado)Image should contain the label 'Avocado' with a confidence level of 90% or higher.