Founder & CEO of MiShen Technology
PhD graduate from the Multimedia Laboratory (MMLab) at The Chinese University of Hong Kong, with long-term research in multimodal generation, 3D reconstruction, and human preference modeling. Has published multiple research achievements at top international conferences such as CVPR, ICCV, and ECCV, and proposed systematic human visual preference modeling work (HPS series), providing important references for preference alignment and evaluation of generative models. Previously conducted cutting-edge AI research and engineering practices at SenseTime, Meta, and other institutions, participating in the development and implementation of multiple generative AI systems for real business scenarios. In recent years, he has focused on the structural connection between AI and real user needs, proposing the Vibe User Research concept, exploring a new generation of product self-iteration and self-upgrade paradigms centered on continuous user feedback and preference modeling, committed to promoting the large-scale implementation of open source technology in enterprise-level AI applications and business innovation.
