You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently I am searching more about this metrics.
Using the IoU distance we obtain a value for MOTP, for example 0.8. What that means?
What is the range for the MOTP metric? And what is the optimal MOTP value?
For what is written in the github:
Metric MOTP seems to be off. To convert compute (1 - MOTP) * 100. MOTChallenge benchmarks compute MOTP as percentage, while py-motmetrics sticks to the original definition of average distance over number of assigned objects
So do we need to make that conversion to get the right value? In my example it would be 20%. Is it a good value?
The text was updated successfully, but these errors were encountered:
Currently I am searching more about this metrics.
Using the IoU distance we obtain a value for MOTP, for example 0.8. What that means?
What is the range for the MOTP metric? And what is the optimal MOTP value?
For what is written in the github:
So do we need to make that conversion to get the right value? In my example it would be 20%. Is it a good value?
The text was updated successfully, but these errors were encountered: