Open zosel260 opened 1 year ago
Hello. Thank your for sharing the wonderful works.
I tested MOT20 with the pre-trained model by following athor's instructions below:
python3 tools/track.py -f exps/example/mot/yolox_x_mix_mot20_ch.py -c pretrained/bytetrack_x_mot20.pth.tar -b 1 -d 1 --fp16 --fuse --match_thresh 0.7 --mot20
After test has done, the terminal shows messages like:
WARNING __main__"122 - No ground thruth for MOT20-04, skipping ...
and The printed the erformance metrics like RcLL, PRCN, FP, FN, IDs,.... are all 'NaN'
I understand this is because MOT20 test dataset doesn't have gt/gt.txt .
I wonder this is normal or abnormal(so I have to fix any problem).
Is there anyway to get gt.txt for MOT test dataset?
Or do I have to use only the offical evaluation toolkit offered by the MOT challenge ?
Hello. Thank your for sharing the wonderful works.
I tested MOT20 with the pre-trained model by following athor's instructions below:
python3 tools/track.py -f exps/example/mot/yolox_x_mix_mot20_ch.py -c pretrained/bytetrack_x_mot20.pth.tar -b 1 -d 1 --fp16 --fuse --match_thresh 0.7 --mot20After test has done, the terminal shows messages like:
WARNING __main__"122 - No ground thruth for MOT20-04, skipping ...and The printed the erformance metrics like RcLL, PRCN, FP, FN, IDs,.... are all 'NaN'
I understand this is because MOT20 test dataset doesn't have gt/gt.txt .
I wonder this is normal or abnormal(so I have to fix any problem).
Is there anyway to get gt.txt for MOT test dataset?
Or do I have to use only the offical evaluation toolkit offered by the MOT challenge ?