Closed acmff22 closed 4 years ago
Hi @acmff22 , I use https://github.com/thodan/bop_toolkit to calculate the metrics, where Average Recall is used to report the results.
The difference of AR and AP can be found here (https://en.wikipedia.org/wiki/Precision_and_recall). Apparently, AR is more suitable than AP to know how well the model detects an object.
Thanks for sharing your great work!!! In your paper, you present the results for the Average Recall of ADD metric, I wonder how the AP calculated? And what differences between the average precision and average recall in this case?