Teaching mode, as the core element of artistic expression, directly affects the aesthetic expression and emotional transmission of students’ works. Based on the application of style migration algorithm supported by artificial intelligence in teaching, this paper analyzes its effect on the innovation of teaching mode and the enhancement of artistic expression. The research adopts a mixed method of technology development and effect evaluation, selects a variety of film and television works with distinctive performance characteristics, and uses convolutional neural networks to extract their performance feature vectors and construct a style database. A style migration system based on generative adversarial network is developed to stylize the new performance materials. The effect of the algorithm is evaluated in three dimensions: content consistency, emotional expression accuracy, and style recognition through expert evaluation and audience perception experiments. It was found that the expert score of the style migration algorithm’s target performance style characteristics was (9.1±1.4) points, and the audience perception of the integration of personal style and target style reached (8.8±0.9) points. It is said that the technology not only enhances the teaching efficiency, but also provides new possibilities for students’ artistic expression ability. The study shows that algorithmic film and television acting teaching is a powerful tool to strengthen the teaching mode, provide theoretical support and practical reference for enhancing artistic performance, and effectively promote the deep integration of film and television acting teaching and artificial intelligence technology.
