Beginners need a lot of practice to master fingering as a motor skill. However, in the post-lesson practice sessions, beginners often fail to provide internal feedback due to a lack of proper long-term memory, which ultimately leads to high error rates and poor practice results. This paper proposes a system to assist in the practice of Guzheng fingering. By constructing three corresponding student types and feedback implementation scenarios to provide different types of augmented feedback, the system aims to better support beginners in practicing Guzheng fingering on their own. The system employs gesture recognition technology to support the feedback intervention, and positive results were obtained through comparative experiments in which the system’s recognition and judgement functions could be used.