Introduction
Mean average precision (mAP) has been widely used in object detection and image retrieval tasks, especially the person re-identification task. Here I will illustrate the calculation of mAP for person re-ID.
Calculation
Briefly, average precision (AP) is the area under the precision-recall curve. In person re-ID, the gap between every recall value can be calculated as recall - old_recall = 1 / num_of_true_matches. Then the area is easy to get. Here is the Python sample code.
for i in range(num_of_good):
# The gap between every recall value
d_recall = 1.0 / num_of_good
precision = (i + 1)*1.0 / (rows_good[i] + 1)
if rows_good[i] != 0:
# The last precision, so i = i + 1 - 1
old_precision = i * 1.0 / rows_good[i]
else:
# Avoid zero
old_precision = 1.0
# ap is the area under the p-r curve
ap = ap + d_recall * (old_precision + precision) / 2
And the mAP is just the average over all query images.
