Wuyang Li

Papers

Total Citations

H-Index

About

Wuyang Li is at the forefront of applying generative AI and advanced 3D/4D reconstruction to medical imaging and surgical robotics. His work bridges the gap between computer vision and clinical practice, with a particular focus on creating high-fidelity, dynamic scene representations for endoscopy. Li's most impactful contribution is **Endora** (2024, 52 citations), a pioneering framework that repurposes video generation models as realistic endoscopy simulators, offering a scalable solution for surgical training without requiring real patient data. He further advanced surgical scene understanding with **LGS** (2024, 42 citations), introducing a light-weight 4D Gaussian Splatting method that enables efficient, real-time reconstruction of deformable tissues during procedures. Beyond medicine, Li addresses critical issues in AI security with his latest work, **Hide-in-Motion** (2025), which embeds steganographic copyright information into 4D Gaussian Splatting assets—a vital step for protecting intellectual property in dynamic scene generation. By combining high-impact methodological innovation with pressing real-world applications, Li is shaping the future of trustworthy, efficient, and clinically deployable AI in surgery and beyond.