关键词: 眼动跟踪, 乐谱视唱, 头部运动偏移, 注视精度优化, Vision Transformer模型

Abstract:

To address the fixation offset problem caused by head movement in music solfeggio teaching simulation and the lack of system-level simulation validation in existing methods, this paper proposed a fixation accuracy optimization method integrating image semantic understanding, temporal trajectory modeling, and solfeggio cognitive simulation. With Vision Transformer as the core, after preprocessing via Mahalanobis distance, sliding window, and region of interest, position offset perception, offset residual regression, and dual-pathway fusion were introduced to achieve offset modeling and correction under unlabeled conditions. Simulation results indicate that the error of this method decreases by 43.9% compared with the original value error; removing any module significantly increases the average Euclidean distance, with a maximum increase of 48.6%; in cross-dataset experiments, the correction rates across different datasets remain at around 40%; the offset error is reduced by 36.6%~40.9% on average in different task scenarios. This method improves the reliability of eye tracking data and provides technical support for solfeggio cognitive assessment and human-computer interaction simulation systems.

Key words: eye tracking, music solfeggio, head movement offset, fixation accuracy optimization, Vision Transformer model