Skip to content
ml-recall

#panoptic-quality

1 question from Segmentation

Timed test

1 question · 1 min · score and explanations at the end

Take a test
Q 1 / 1Score 0 / 0

Question 1: Short answer

hardAI-checkedSegmentation

For the class "person", an image has three ground-truth segments and the prediction has three segments. Two predicted segments overlap ground-truth persons with IoU 0.9 and 0.6; the third overlaps the remaining ground-truth person with IoU 0.4. A predicted and a ground-truth segment match only if their IoU is strictly greater than 0.5. Panoptic quality is

PQ=∑(p,g)∈TPIoU(p,g)∣TP∣+12∣FP∣+12∣FN∣.\text{PQ} = \frac{\sum_{(p, g) \in TP} \text{IoU}(p, g)}{|TP| + \tfrac12 |FP| + \tfrac12 |FN|}.

What is PQ for this class?

a number between 0 and 1