The present application relates to the technical field of artificial intelligence. Disclosed are an interview method and apparatus based on an AI interview system, and a computer device and a storage medium. The method comprises: outputting a preset first question to a user end; receiving a first answer of the user end replying to the first question, and obtaining an intention of the user end; collecting a facial image of a user corresponding to the user end, and inputting the facial image into a preset subtle expression recognition model obtained through training on the basis of a neural network model and calculating same, so as to obtain a nervous emotion value of the user; if the nervous emotion value of the user is within a preset emotion range value, acquiring a second question; outputting the second question to the user end; receiving a second answer of the user end replying to the second question; and if a preset interview termination trigger condition is met, generating an interview report. A questionnaire in which flexible jumping is enabled is provided, and the quality of questionnaire answers is guaranteed; and in addition, an abnormal interviewee can be eliminated.
展开▼
