Research Projects

Recent advances in conversational artificial intelligence and real-time avatar technologies are creating new opportunities for more accessible and interactive digital healthcare services. The Real-Time Avatar-Based Smart Medical Interview System combines intelligent dialogue processing, natural language understanding, and interactive avatar interfaces to provide patients with a more intuitive and human-centered medical interview experience.

Rather than relying on conventional static questionnaires, the system enables users to communicate naturally with a virtual medical assistant that dynamically responds to their answers and guides them through an adaptive consultation process. This conversational approach allows the system to gather medical information in a structured yet flexible manner while improving accessibility and user engagement.

From a technical perspective, the system integrates large language models, speech processing, real-time avatar rendering, and context-aware dialogue management to analyze patient responses and generate appropriate follow-up questions.

Collected information can be transformed into structured clinical summaries, enabling healthcare professionals to review essential patient information more efficiently. The platform aims to reduce repetitive preliminary consultation workloads while improving the consistency of patient information collection.

Recent developments in educational technology have established the AI-driven educational solution framework as a transformative paradigm for modern learning environments. Our platform combines adaptive artificial intelligence, generative AI, and real-time learning analytics to provide educational experiences tailored to each learner's proficiency, progress, and individual learning needs.

By moving beyond conventional one-size-fits-all curricula, the system constructs dynamic learning pathways that continuously adapt to student performance. Student interactions and assessment results can be analyzed to identify knowledge gaps, recommend appropriate resources, generate customized exercises, and adjust the difficulty and structure of educational content.

The educational architecture integrates large language models, intelligent recommendation engines, learning analytics, retrieval-augmented generation, and automated content generation . These technologies support personalized learning while providing educators with analytical information that can be used for targeted interventions and more efficient instructional management.