You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
mAP (Mean Average Precision) is a metric to measure the accuracy of various object detection models. You can measure mAP for different IoU thresholds (eg: 0.5, 0.7 0.9).
Steps to calculate maP:
Calculate precision and recall for all the bounding boxes(with different IoU).
Use the point interpolation method to plot the graph of recall versus precision.
Take the average of all interpolation values to get the Average Precision.
Take the mean of all AP with different IoU. The result will be the final mAP.
for what measure_map specifically used??
The text was updated successfully, but these errors were encountered: