Python: Advanced Guide to Artificial Intelligence

The mean average precision

The mAP is used for evaluating detection algorithms. The mAP metric is the product of precision and recall of the detected bounding boxes. The mAP value ranges from 0 to 100. The higher the number, the better it is. The mAP can be computed by calculating average precision (AP) separately for each class, then the average over the class. A detection is considered a true positive only if the mAP is above 0.5.  All detections from the test images can be combined by drawing a draw precision/recall curve for each class. The final area under the curve can be used for the comparison of algorithms. The mAP is a good measure of the sensitivity of the network while not raising many false alarms. We have learned the evaluating ...