Determining a Person's Membership in the Music Community with VGG-Face and Vision Transformer

The use of machine learning in pathognomy is being investigated to test the hypothesis about the influence of classical music education on a person's face. A sample of images of the faces of symphony orchestra musicians and people without musical education was collected, which is used to solve the binary classification problem. The quality of the sample is analyzed using dimensionality reduction algorithms. The pre-trained VGG-Face model and a Vision Transformer are used to obtain embeddings. The values of the quality metrics in the test sample allow us to state with some degree of confidence that it is possible to determine whether a person is a musician of a symphony orchestra based on the image of his or her face.