About yolov3, conduct map test directly

Many blogs use valid Function in two steps for various AP To test , And then calculate mAP, For specific steps, please refer to various blogs ,
Actually in darknet The model after training can be directly mAP The calculation of ,
stay readme It says :

To check accuracy [email protected]=50: `darknet.exe detector map data/obj.data yolo-obj.cfg backup\yolo-obj_7000.weights`
# That is, the model can be calculated directly with the above command mAP

That is, the model can be calculated directly with the above command mAP, And IoU=0.5, Of course,

To check accuracy [email protected]=75: `darknet.exe detector map data/obj.data yolo-obj.cfg backup\yolo-obj_7000.weights -iou_thresh 0.75`
# You can get... Directly IoU=0.7 Of mAP

another :
Not after training xx_final.weights Of mAP The highest , By command

* `darknet.exe detector map data/obj.data yolo-obj.cfg backup\yolo-obj_7000.weights`
* `darknet.exe detector map data/obj.data yolo-obj.cfg backup\yolo-obj_8000.weights`
* `darknet.exe detector map data/obj.data yolo-obj.cfg backup\yolo-obj_9000.weights`

To detect and final.weights Compare , Get the best model .