A sawmill knot classifier must keep its agreed error band when 12% of frames are motion-blurred, when a share of labels in a dirty feed are skewed, and when the mill lights flicker for half a minute — it may degrade to a slower, coarser mode, but it must not crash or silently abandon the band. What acceptance criterion is that?
Select an answer to reveal the explanation.
Short Explanation
Keep the agreed error band when 12% of frames are motion-blurred, labels are skewed, and mill lights flicker. A slower, coarser mode is allowed; a crash or a silent abandon of the band is not. That is AI robustness acceptance. Flicker is not a retarget time.
Full Explanation
Robustness acceptance keeps the agreed correctness band under hostile or broken inputs. It may allow a slower, coarser mode. Adaptability, transparency, and a binary ban on degradation miss that example.