Enhancing E-Learning Accessibility with AI-Driven Multilingual and Complexity-Adaptive Subtitle Generation and Translation

Abstract - Enhancing the accessibility of e-learning sites is critical to a broad spectrum of users. The spread of content is constrained by the fact that there is no automation and multilingual support in classic subtitle creation. Relying on the assistance of speech-to-text technology, this research makes a contribution of an AI-facilitated method to automatic subtitle creation and translation such that English subtitles could be created effortlessly for learning videos. The titles are also translated into Malayalam and Tamil by AI translation for convenience to the learners in the region. The system also offers dual-level subtitles, with simplified English for children and beginners while keeping complex English for advanced learners. This makes the content accessible to different audiences with different levels of language proficiency. The system preserves context meaning and quality of translation using improved natural language processing capabilities. This facilitates wider learning online with the use of multilingual translation in combination with automatic transcriptioning. More consumers can access e-learning content depending on benchmark testing showing a big improvement in precision, usability, and flexibility of use to serve various user purposes.