Hello.
I found that in our file "gt_label_ucf.npy" , the label for normal videos are set to label "A", but when calculating the mAP in file "ucf_detectionMAP.py" , we are still matching samples with the label "Normal". This seems to cause some incorrect results in the mAP calculation.
Hello. I found that in our file "gt_label_ucf.npy" , the label for normal videos are set to label "A", but when calculating the mAP in file "ucf_detectionMAP.py" , we are still matching samples with the label "Normal". This seems to cause some incorrect results in the mAP calculation.