plotting the result of a detector problem

I want to evaluate a multiclass detector, but the propblem is I didn't get a suitable plotting results. I need the average result of the detector as one number, not a number for each class.

In the following the code I used, if there is anything wrong mention it, please.

threshold =0.5;

[averagePrecision, recall, precision] = evaluateDetectionPrecision(results, TestData(:,2:end),threshold);

[logAverageMissRate,fppi,missRate] = evaluateDetectionMissRate(results, TestData(:,2:end),threshold);

Plot the evaluation metrics

subplot(1,2,1);
plot(recall{1,1},precision{1,1},'g-','LineWidth',2, "DisplayName",'1st');
hold off;
xlabel('Recall');
ylabel('Precision');
title(sprintf('Average Precision = %.2f\n', averagePrecision))
legend('Location', 'best');
legend('boxoff')
grid off

subplot(1,2,2);
plot(fppi{1,1}, missRate{1,1},'-g','LineWidth',2, "DisplayName",'2nd');
hold off;
xlabel('False Positives Per Image');
ylabel('Log Average Miss Rate');
title(sprintf('Log Average Miss Rate = %.2f\n', logAverageMissRate))
legend('Location', 'best');
legend('boxoff')
grid on

Answer 1

[Srivardhan Gadila]

As you have already mentioned that you are performing multi-class object detection & based on the above plots I think that there are four classes i.e., four different object classes to detect.

For a multiclass detector, As per the documentation of evaluateDetectionPrecision,

1. The average precision is a vector of average precision scores for each object class
2. The recall and precision are cell arrays, where each cell contains the data points for each object class.

And as per the documentation of evaluateDetectionMissRate,

1. The log-average miss rate is returned as a vector of values that correspond to the data points for each class.
2. The FPPI and log miss rate are cell arrays, where each cell contains the data points for each object class.

Hence you need to make use of the mean function and take the average along appropriate dimension for each of the above outputs.