The development of bio-sensing technology provides new technical support for classroom teaching state recognition and collaborative process regulation. Focusing on the problems of feedback lag, insufficient state judgment accuracy and insufficient collaborative adjustment in teacher-student collaborative teaching, this paper constructs an optimization framework of teacher-student collaborative teaching based on biological perception technology, and integrates heart rate, heart rate variability, electroskin response, EEG attention features, classroom behavior logs and learning platform data into a unified analysis process. A total of 124 students from two natural classes were selected to carry out the 12-week quasi-experiment, including 62 students in the experimental group and 62 students in the control group. The results showed that the final test score of the experimental group was 86.37, higher than that of the control group (81.14). The score of classroom project was 88.52, higher than 82.63 of the control group; The learning satisfaction was 4.46. Research shows that this model can improve the accuracy of classroom state recognition, optimize teachers ‘feedback path, and promote the transformation of teacher-student collaborative teaching from experience regulation to data-driven improvement.
