Analysis of Ear Canal Echograms for Personality Verification Using AIConstructor Software Package

Biometrics is one of the responsible applications of artificial intelligence (AI). The most important tasks today are to increase the reliability of biometric authentication and protect biometric images from compromise. The article is devoted to the study of ear canal echograms for the purpose of their use for user authentication of computer systems using the AIC-ears-75, which is an open dataset. The AIC desktop software package developed by the authors was used as a research tool. An autoencoder based on convolutional neural networks was used to extract features of the ear canal data. Ear images were verified using сlassifiers on the basis of formulas of Bayes hypothesis, a shallow neural network, and an artificial immune network used to classify ear images. The best result was shown by using the improved Bayesian classifier: EER≈0.08.