Аннотация:The purpose of this study is to study the phenomenon of emotional mirroring using remotely diagnosed dynamic parameters of facial expressions. The main characteristic of mirroring is the subconscious adjustment and copying of the dynamics of another person, which allows you to create a subtle, elusive connection, contact between people. We explored reflection as a reproduction of the emotions of one person in a telecommunication dialogue through a cognitive-emotional conflict. We developed a computer program using VGG16-based neural network algorithms to mark people’s emotional reactions in video data automatically. We used recordings of 24 interviews with the participants of both genders (10 males and 14 females between the ages of 12 and 27). Three qualified experts-psychologists participated in the study; a total of 338,075 frames of correct data were obtained. We used interviews based on self-attitude techniques, problematic interviews based on transactional analysis, free reasoning about controversial and topicalsituations,thesituationswereselectedconsideringtheageandotherindicatorsofthegroupofparticipants, and watching video clips with a surprise ending. As a result, it was found that the parameters of facial expressions of the participant and the experimenter (psychologist) allocated by the program strongly correlate with such indicators as happy, sad, and surprise. Strong negative correlations are found between the parameters of the happiness of participant and fear of psychologist, sad of the participant and happiness of the psychologist, sad of psychologist and surprise of the participant, and a direct relationshipbetween sad of participant and fear of psychologist. Two general factors describe the revealed correlations. Simultaneously, all ofthe identified correlations appear both in the situation with and without cognitive-emotional conflict, but their expression is firmlyand reliably different in these situations.