QuestionQ77

VerifiedMonitoring AI solutions

You recently developed the first version of an image-segmentation model for a self-driving car. After deploying it, you notice a decline in the area-under-the-curve (AUC) metric. While reviewing video recordings, you also find that the model fails in highly congested traffic but works as expected when traffic is lighter. What is the most likely cause of this outcome?

Explanation

Condition-specific performance in which a model succeeds in less-complex traffic scenes but fails in highly congested scenes is consistent with overfitting to lower-traffic patterns and inadequate fitting or generalization for the more complex congested-traffic distribution.

Community Discussion

No comments yet. Be the first to start the discussion!