基于LLM和CV的AI个人健身教练系统设计与实现  

A Personalized AI Fitness Coach System Powered by LLM and CV

作　　者：刘卓鑫 冯月春[1] LIU Zhuoxin;FENG Yuechun(Ningxia Institute of Science and Technology,Shizuishan 753000,China)

机构地区：[1]宁夏理工学院,宁夏石嘴山753000

出　　处：《河南科技》2025年第2期24-28,共5页Henan Science and Technology

基　　金：宁夏回族自治区2023年省级大学生创新项目(S202312544008)。

摘　　要：【目的】为了给用户提供个性化健身指导,设计出一种基于LLM大语言模型和计算机视觉(CV)的AI个人健身教练系统。【方法】采用PP-TinyPose进行关键点匹配,并使用微调后的InternLM2大语言模型,使系统能通过摄像头捕捉用户运动姿态并进行分析,从而提供个性化的健身指导。【结果】测试结果表明,该系统能准确识别运动姿态,并提供实时反馈和建议,显著提高健身效果。【结论】基于PP-TinyPose和微调后的InternLM2模型的AI健身教练系统能为用户提供有效的个性化健身指导,具有广泛的应用前景。[Purposes]To provide personalized fitness guidance to users,an artificial intelligence(AI)personal fitness coach system based on a large language model(LLM)and computer vision(CV)is proposed.[Methods]The system utilizes PP-TinyPose for key point detection and a fine-tuned InternLM2 large language model.By capturing and analyzing users'exercise postures through the camera,the system generates personalized fitness guidance.[Findings]The test results demonstrate that the system can accu⁃rately recognize exercise postures and provide real-time feedback and suggestions,significantly improving fitness outcomes.[Conclusions]The AI fitness coach system proposed in this paper,based on PP-TinyPose and the finetuned InternLM2 model,effectively provides users with personalized fitness guidance and shows great potential for wide application.

关 键 词：人工智能 PP-TinyPose XTuner InternLM2 个人健身教练 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP391.41[自动化与计算机技术—控制科学与工程]