<![CDATA[The integration of artificial intelligence and deep learning into physical education (PE) has created new opportunities to enhance students' self-regulated learning (SRL), yet evidence on the effectiveness of deep learning-based learning media remains fragmented. This systematic literature review aimed to synthesize existing evidence on the application of deep learning-based learning media and its impact on students' SRL in PE. Following the PRISMA 2020 guidelines, a systematic search of the Scopus database identified studies published between 2015 and 2025. After a multi-stage screening process based on predefined eligibility criteria, 28 studies were included and analyzed using thematic synthesis. The review identified three major findings: deep learning-based learning media enhanced students' goal setting and self-monitoring through computer vision, intelligent tutoring, and adaptive feedback systems; multimodal feedback improved metacognitive awareness and motor skill acquisition; and considerable heterogeneity in study designs and SRL measurement instruments limited cross-study comparisons. Overall, deep learning-based learning media show strong potential to promote SRL in PE by supporting cognitive, metacognitive, and motivational learning processes. Future research should prioritize longitudinal designs, standardized SRL assessment instruments, and validation across diverse educational contexts.]]>
Copyrights © 2026
