Calculating the mAP?

[SWRIGH211]

Hi there,
I'm trying to compare my custom trained yolo5s model against other models, however they do no state the mAP@.5 or the mAP@.5:.95. only AP. My question is, how do you attain the AP of your model. I had a look at this medium explanation:

https://medium.com/@timothycarlen/understanding-the-map-evaluation-metric-for-object-detection-a07fe6962cf3

However im still quite confused on how to calculate it properly.

Does anyone have a clearer simpler explanation, it would be much appreciated.

Many thanks in advance

[SWRIGH211]

its in relation to the The mean Average Precision (mAP) quantitative measure from PASCAL VOC 2010.

[glenn-jocher]

@SWRIGH211 VOC primary metric is mAP@0.5.

mAP computation code is here:

yolov5/utils/metrics.py

Lines 18 to 31 in ed2c742

	def ap_per_class(tp, conf, pred_cls, target_cls, plot=False, save_dir='.', names=()):
	""" Compute the average precision, given the recall and precision curves.
	Source: https://github.com/rafaelpadilla/Object-Detection-Metrics.
	# Arguments
	tp: True positives (nparray, nx1 or nx10).
	conf: Objectness value from 0-1 (nparray).
	pred_cls: Predicted object classes (nparray).
	target_cls: True object classes (nparray).
	plot: Plot precision-recall curve at mAP@0.5
	save_dir: Plot save directory
	# Returns
	The average precision as computed in py-faster-rcnn.
	"""

[SWRIGH211]

thankyou makes sense