Open amoskalev opened 4 years ago
AUC (Area-Under-Curve) is calculated by separating the curve into several discrete bins (nbins=50
in the code), taking each bin as a rectangle, and sum over the areas of these bins. In formular:
AUC = sum_i^nbins f(xi) * d_xi
where d_xi = 1 / nbins
. So AUC is the mean over f(x_i)
s.
The precision score is defined as the percentage of successfully tracked frames under the criterion: center error <= 20
(see OTB benchmark paper for details), which is prec_curve[20]
.
Hi, my question is about how OTB experiment results are calculated, i.e. success and precision scores.
In the OTB benchmark, areas under success and precision curves are used, while in your code I see that you take the mean to calculate
succ_score
. Same withprec_score
, I don't understand why do you calculate them like that.Here is how metrics are calculated in ExperimentOTB:
Shouldn't you take AUC instead of the mean value?