A Mobile Multimodal Interactive System for Sports Skill Assessment in Educational Contexts

Abstract

Traditional sports skill assessment has been constrained by subjective bias, delayed feedback, and limited evaluation dimensions, thereby restricting improvements in instructional quality. The integration of mobile technologies and multimodal interaction has introduced new opportunities for precise skill assessment; however, existing systems commonly suffer from inadequate mobile adaptability, inefficient multimodal data collaboration, and limited suitability for authentic teaching scenarios. In this study, a mobile multimodal interactive skill assessment system tailored for sports education was designed to enable real-time, accurate, and multidimensional evaluation of sports skills in support of instructional optimization. A four-layer architecture comprising mobile terminals, data transmission, cloud-based processing, and application services was established. Multimodal data acquisition modules were integrated with lightweight fusion algorithms, and system effectiveness was examined through requirement analysis, architectural design, prototype development, and teaching-oriented experimental validation. The results demonstrate that the proposed system exhibits robust stability in mobile environments and achieves high multimodal data fusion accuracy. Compared with traditional assessment approaches and existing systems, assessment accuracy is significantly improved, and reliable application performance is observed in practical sports teaching contexts. The primary contribution lies in the development of a lightweight multimodal acquisition and real-time assessment framework specifically adapted to sports education scenarios. This work enriches the theoretical foundation of mobile intelligent assessment and provides critical technical support for advancing the intelligent transformation and scientific assessment of sports teaching.