The image is being processed, do not close the tab or the browser until it is finished, thanks.


Legal age estimation using

This is an example of an image used to train our model. The model was trained with data from 12 countries and 4 continents so it can generalize to data from never seen samples. It obtained a mean absolute error of 1.07 years and an accuracy of 88.38% (for the 18 years threshold) in test. Regarding never seen samples, it obtained a mean absolute error of 1.21, 1.45, 1.36 and 1.51 years and an accuracy of 92.04, 81.97, 85.03 and 86.08% in four samples from Russia, Ethiopia, Egypt and Australia, respectively.

Coverage level PPV * Sensitivity Specificity Accuracy
0 (original model) 0.888 0.923 0.824 0.883
60% 0.952 0.811 0.939 0.861
80% 0.970 0.707 0.967 0.810
90% 0.982 0.620 0.982 0.764
95% 0.983 0.541 0.986 0.717
99% 0.993 0.318 0.996 0.587

* Positive predictive value

These are the results of our model when estimating age as the minimum of aprediction interval for different coverage levels. The coverage level determines the percentage of the sample used to obtain the maximum error that defines the limits of the interval, excluding the outermost value (potential outliers).


Remember to use an image like the one in the example for correct operation of the tool. (Supported format .jpeg).

Check this box if you want to get a heat map of what the AI is identifying (will increase estimation time considerably).
I accept the Terms and Conditions.