In computer vision and object detection, the common evaluation method is mAP. What is it and how is it calculated?
Definition: mAP → mean Average Precision
In most of the object detection contests, there are many categories to detect, and the evaluation of the model is performed on one specific category each time, the eval result is the AP of that category.
When every category is evaluated, the mean of all APs is calculated as the final result of the model, which is mAP.