false postive in MOTA

[miketry0313]

Hello,

Thanks for sharing your code.

When the total number of hypothesis is much greater than the number of ground truth, is it possible that false positive is too large to make MOTA < 0?

Thanks.

Mike

[DenisKuplyakov]

@miketry0313 Yes, it's possible. The only thing you can say is that MOTA <= 1.0.

[EddieEduardo]

Hi，I did encountered the same problem here, my result of MOTA/MOTAL is negative， which means my tracker performance is extremely bad？
Thanks for reply.