Summary
Xiaochuan Fan is a Machine Learning Engineer with eight years of experience applying computer vision and multi-modal ML to real-world products, currently shaping search relevance and content understanding for TikTok Search US. He previously led multi-modal search, image recognition, and AR/VR initiatives at JD.com and developed efficient detection and low-shot learning methods for HD mapping and autonomous driving at HERE. A Ph.D. in Computer Vision from the University of South Carolina, he has published over ten papers in top venues (CVPR, ICCV, ECCV, KDD, IJCV, SIGIR), bridging deep research with production systems. Known for practical innovations—such as dual-source CNNs for pose estimation and data-driven 3D pose models—he combines rigorous academic foundations with hands-on delivery of scalable, multimodal search and recommendation features. Based in Mountain View, he brings domain depth in short-video and e-commerce search plus a knack for turning cutting-edge research into product impact.
8 years of coding experience
6 years of employment as a software developer
Doctor of Philosophy (Ph.D.) Computer Science and Engineering Computer Vision, Doctor of Philosophy (Ph.D.) Computer Science and Engineering Computer Vision at University of South Carolina
English, Chinese