1
$\begingroup$

http://cocodataset.org/#detection-eval "ARmax=1: AR given 1 detection per image" "AR is the maximum recall given a fixed number of detections per image, averaged over categories and IoUs."

Above link, they described ARmax=1, but I didn't understand exactly. I understood general recall concept in machine learning or other areas. Could you explain ARmax=1 more detail ??

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

AR is the abbreviation for Average Recall.

max=1 only looks for 1 detection per image.

Thus, ARmax=1 is the average ability of the model to find a given category in an image. For example if there is a dog in the image, the model should predict a dog in the image. That performance is averaged across all categories.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.