ImageSegmentationEvaluationMetrics

Metrics for image segmentation evaluation results.

Fields
confidenceMetricsEntries[] object (ConfidenceMetricsEntry)

Metrics for each confidenceThreshold in 0.00,0.05,0.10,...,0.95,0.96,0.97,0.98,0.99 Precision-recall curve can be derived from it.

JSON representation 
{
  "confidenceMetricsEntries": [
    {
      object (ConfidenceMetricsEntry)
    }
  ]
}

ConfidenceMetricsEntry

Fields
confusionMatrix object (ConfusionMatrix)

Confusion matrix for the given confidence threshold.

confidenceThreshold number

Metrics are computed with an assumption that the model never returns predictions with score lower than this value.

recall number

Recall (True Positive Rate) for the given confidence threshold.

precision number

Precision for the given confidence threshold.

diceScoreCoefficient number

DSC or the F1 score, The harmonic mean of recall and precision.

iouScore number

The intersection-over-union score. The measure of overlap of the annotation's category mask with ground truth category mask on the DataItem.

JSON representation 
{
  "confusionMatrix": {
    object (ConfusionMatrix)
  },
  "confidenceThreshold": number,
  "recall": number,
  "precision": number,
  "diceScoreCoefficient": number,
  "iouScore": number
}