“Multimodal Attention-Based Mixture of Experts for Video-Level Activity Recognition using the YouTube-8M Dataset” (2026) ITEGAM-JETIA, 12(61), pp. 1490–1499. doi:10.5935/jetia.v12i61.4267.