Jinyuan Xu 徐瑾源

HKUST (Guangzhou)

Master’s in Artificial Intelligence

Anhui University

Bachelor’s in Statistics

University scholarship (top 10%); provincial undergraduate innovation project.

vivo · Research Intern

Imaging Algorithm Research Department

Draft-and-Render: Mixed-Resolution Distillation for Video Super-Resolution

Sole first author / Ongoing · Targeting CVPR

XLeaF Lab · Research Member

Research on visual generative models and image-to-video generation.

Golden Anchor for Video Generation

Jinyuan Xu, Shitong Shao, Lichen Bai, Zikai Zhou, Guoxi Huang, Nantheera Anantrasirichai, Zeke Xie

• First-author research on learning anchor images for downstream video generation.
• Trains a text-to-image generator using video-level rewards and offline preference optimization.
• Studies transfer across four image-to-video models.

星火AI动漫

Founder and creator of AI-generated visual stories and short-form video.