The proliferation of artificial intelligence (AI) in schools like higher ed has largely been centered around learning through text-based sources, often lacking visual components, even though these are essential in many subjects (STEM): science, technology, engineering, and mathematics. In this paper, we propose an AI-focused learning framework that integrates Microsoft Copilot Vision to accelerate visual perception and understanding in university settings. The system will interpret pedagogical diagrams, charts, and handwritten notes to develop them into organized textual explanations and interactive learning materials. It aims at making text with natural language generation accessible and engaging as well as encouraging cognitive development in learners with image recognition. Experimental observation on open education datasets and real classroom resources show to better enhance learner understanding and lower the cognitive load than classical static visual aids. This study provides a foundation for intelligent, multimodal learning that closes the gap between visual and linguistic perception, leading to personalized and inclusive education in higher learning institutions.