search

AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
http://pjreddie.com/darknet/
Other
21.74k stars 7.96k forks source link

detector.c validate_detector_map #2579

Open weiting0032 opened 5 years ago

weiting0032 commented 5 years ago

image

@AlexeyAB I want to ask why avg_precision neet to divide by 11 and the for loop point need to be pass 12 times?

AlexeyAB commented 5 years ago

@WEITINGLIN32

and the for loop point need to be pass 12 times?

The body of loop for(point = 0; point < 11; point++) will be executed only 11 times (not 12): https://ideone.com/ir2c1M

weiting0032 commented 5 years ago

@WEITINGLIN32

and the for loop point need to be pass 12 times?

The body of loop for(point = 0; point < 11; point++) will be executed only 11 times: https://ideone.com/ir2c1M

Why it is 11 times?

AlexeyAB commented 5 years ago

@WEITINGLIN32 Because we should get 11 points from Precision-Recall curve: https://medium.com/@jonathan_hui/map-mean-average-precision-for-object-detection-45c121a31173