Sort metrics by the question they answer

Sort each metric into the question it best answers. Overall correctness Alert trust or catch rate Localization quality Workflow speed Accuracy: what share of all image classifications were correct? Precision: when the model alerts, how often is it right? Recall: how many true cases did the model catch? IoU: how much does the predicted box or mask overlap the ground truth? mAP: how well does object detection perform across classes and thresholds? Latency: how long does the model take to return a prediction? False positives per hour: how much review workload will alerts create? Frames per second: can the system…

Sign up free — one personalized lesson every day, matched to your role and goals.

Already have an account? Sign in