← Return

His algorithm aims for better age assessments

Anders Hast has developed a method of his own for facial recognition by combining AI and old algorithms developed in the late 20th century.

Anders Hast, a Professor of Image Processing, is developing an algorithm that uses facial recognition technology to determine a person's age based on a photo. Traditional medical methods for assessing age have been criticized for their inaccuracies, and Hast aims to investigate whether facial recognition can provide a more precise assessment. He has previously used similar algorithms to categorize historical portraits by biological sex, achieving a higher accuracy rate than human assessors. Hast plans to combine modern AI and neural networks with his own algorithm to train an AI tool to recognize typical characteristics of different ages. The algorithm focuses on facial symmetries rather than superficial attributes like hair or beard. The goal is to achieve stable age recognition for different genders and ethnicities. While the method shows promise in recognizing individuals across different ages, accurately determining the actual age remains a challenge. Hast hopes that in the future, the error rate can be reduced to within a year or two in 90% of cases. The research aims to prevent age assessments based on uncertain evidence and contribute to more accurate legal proceedings.