Transfer learning-based adaptive music teaching system for modulating students' emotions

With the widespread application of artificial intelligence technology in the field of education, how to achieve emotion recognition and personalised regulation within the teaching process has become a significant research focus for intelligent teaching systems.This research provides an adaptive music teaching system solution based on transfer learning to address the constraints of traditional music education in emotional perception and delayed feedback.In this framework, learners' multimodal signals are initially acquired synchronously; later, transfer learning is utilised to facilitate emotion recognition and cross-domain feature transfer; ultimately, the system achieves synchronous optimisation of emotional regulation and learning performance.The proposed system shows improvements of 12.4%, 19.1%, 20.3%, and 17.4% in learning performance, engagement, emotional stability, and user satisfaction, respectively, when compared to traditional teaching techniques.This offers innovative theoretical frameworks and technological assistance for the development of emotion-driven intelligent educational systems.