Hi Sahik,
From the example that you are refering to: https://www.mathworks.com/help/vision/ug/object-detection-using-faster-r-cnn-deep-learning.html, we can get the confusion matrix by modifying the section which explains on how to evaluate the object detector using the average precision metric.
classID = 1;
metrics = evaluateObjectDetection(detectionResults,testData);
cf = metrics.ConfusionMatrix % add this line, cf is the confusion matrix
precision = metrics.ClassMetrics.Precision{classID};
recall = metrics.ClassMetrics.Recall{classID};
The function "evaluateObjectDetection" returns an object of type "objectDetectionMetrics" which has a property called "ConfusionMatrix" which returns the confusion matrix. For more information regarding "evaluateObjectDetection" and "objectDetectionMetrics" refer to:
- evaluateObjectDetection: https://www.mathworks.com/help/vision/ref/evaluateobjectdetection.html
- objectDetectionMetrics: https://www.mathworks.com/help/vision/ref/objectdetectionmetrics.html
The confusion matrix obtained for the example stated in the reference is:
Hope this helped and solved your query on how to obtain the confusion matrix!
