<![CDATA[Deaf students in Indonesian special schools (SLB) face persistent barriers in accessing interactive instructional media that align with their visual-spatial cognitive needs and Bisindo linguistic identity. While prior studies have explored sign language recognition technologies and general deep learning algorithms, a critical gap remains in integrating pedagogical deep learning principles with disability-specific interactive media within special education classrooms. This study aimed to develop and validate a Smartboard-based Indonesian Sign Language (Bisindo) deep learning instructional model for deaf students at SLB Pembina Medan. Employing a Research and Development (R&D) approach following the Borg and Gall ten-step model, the study involved needs analysis, product development, expert validation, and field testing. Validation instruments were assessed for content validity using the Content Validity Index (CVI) and Content Validity Ratio (CVR), and reliability was established through inter-rater agreement. The product was evaluated by one learning expert, one Bisindo expert, and one media expert, followed by small-group trials (N = 3 teachers) and large-group trials (N = 3 teachers). Ethical approval was obtained from the institutional review board prior to data collection. Results demonstrated exceptional feasibility: material experts rated the product at 97.06%, Bisindo experts at 95.59%, and media experts at 91.67%, all classified as "very feasible." Small-group teacher trials yielded 88.33% (feasible), while large-group trials achieved 96.66% (very feasible). The developed model offers a replicable pedagogical framework that integrates Bisindo linguistic features, deep learning pedagogy, and Smartboard interactivity, contributing to inclusive digital learning practices for deaf students in special education contexts.]]>
