Abstract Sports participation plays a critical role in holistic education, promoting physical fitness, teamwork, and cognitive development. Traditional methods of monitoring student engagement, such as manual attendance registers or self-reported logs, are often inefficient, error-prone, and lack real-time insights. This study proposes a Deep Learning-Based Image Analytics Framework to automatically monitor and quantify sports participation in academic institutions using computer vision techniques. Leveraging the 100-Sports Image Classification dataset, the framework integrates data preprocessing, augmentation, and state-of-the-art deep learning models including EfficientNetV2, Vision Transformer (ViT), and a hybrid CNN-ViT architecture. Images are standardized, normalized, and augmented to enhance model robustness, while feature extraction combines local spatial and global contextual information to accurately classify sports activities. The models are trained using cross-entropy loss and optimized with AdamW, achieving up to 97.8% test accuracy with the hybrid CNN-ViT model. A Participation Index (PI) is computed based on the number of participants, duration, and activity intensity, enabling quantitative assessment of engagement. Results are visualized through bar charts, line graphs, confusion matrix heatmaps, and pie charts, providing actionable insights for administrators. The framework demonstrates scalability, high accuracy, and real-time monitoring capabilities, addressing limitations of traditional tracking methods. By combining automated classification with participation analytics, this approach provides a reliable, interpretable, and practical solution for academic institutions to monitor student engagement in sports and make data-driven decisions for curriculum and resource planning. Keywords: Sports Participation, Deep Learning, CNN, Vision Transformer, Image Classification, Participation Index, Academic InstitutionsTop of Form Citation M. Anline Rejula, P. Jabalin Reeba “Deep Learning-Based Image Analytics Framework for Monitoring Sports Participation in Academic Institutions” International Journal of Current Science Research (IJCSR) e-ISSN: 2454-5422: 12(4): 2026: 6 - 20
P. Jabalin Reeba M. Anline Rejula (Wed,) studied this question.