Age Verification API Documentation

The Age Verification API utilises a vision-based Artificial Intelligent (AI) model to determine if the age of the input face meets specified criteria. This process involves the following steps:

  1. Estimate Age: The API estimates the age based on the input facial image.
  2. Calculate Distance: The distance is calculated as the difference between the estimated age and a user-defined age threshold.
  3. Determine Confidence: The API uses specific thresholds values to ascertain the confidence level of the age verification result.

Verification Outcomes

For verify_over(image, age_threshold)

verify_over

The distance = estimated_age - age_threshold is the difference between the estimated age and the threshold age.

There are four possible outcomes for the verify_over function:

  • FALSE: the estimated age is less than the threshold age as the distance < 0.
  • TRUE_MODERATE: the estimated age is over the threshold age with moderate confidence, as the distance is between 0 and the High Confidence Distance.
  • TRUE_HIGH: the estimated age is over the threshold age with high confidence, as the distance is between the High Confidence Distance and the Very High Confidence Distance.
  • TRUE_VERY_HIGH: the estimated age is over the threshold age with very high confidence, as the distance is greater than the Very High Confidence Distance.

Note: The High Confidence Distance and Very High Confidence Distance are preset values that determine the confidence level of the age verification result based on our benchmarking on large population datasets. They are 0.75 and 1.5 years respectively for verify_over function.

For verify_under(image, age_threshold)

verify_under

The distance = estimated_age - age_threshold is the difference between the estimated age and the threshold age.

There are four possible outcomes for the verify_under function:

  • FALSE: the estimated age is over the threshold age, as the distance > 0.
  • TRUE_MODERATE: the estimated age is under the threshold age with moderate confidence, as the distance is between 0 and the High Confidence Distance.
  • TRUE_HIGH: the estimated age is under the threshold age with high confidence, as the distance is between the High Confidence Distance and the Very High Confidence Distance.
  • TRUE_VERY_HIGH: the estimated age is under the threshold age with very high confidence, as the distance is smaller than the Very High Confidence Distance.

Note: The High Confidence Distance and Very High Confidence Distance are preset values that determine the confidence level of the age verification result based on our benchmarking on large population datasets. They are -1.3 and -2.6 years respectively for verify_under function.