Summary
Ryan Yuan is a Research Director with 12 years of experience leading computer vision and scene understanding research, currently building a zero-to-one team at Canva focused on next-generation graphic design generation. He spent nearly eight years as a senior researcher at Microsoft, producing top-ranked semantic segmentation methods (OCNet/ISANet/OCRNet, SegFix) with publications at ICCV, ECCV, and IJCV and shipping features into Azure Form Recognizer and Microsoft Designer. His work spans applied research and product impact—from large-scale image search competitions and virtual try-on prototypes to deploying AIGC features—bridging state-of-the-art models with production systems. He holds advanced training in computer vision (PhD-level work at the University of Chinese Academy of Sciences) and earlier degrees from Peking University and Nanjing University. Based in Haidian, Beijing, he actively recruits research interns and full-time researchers and can be reached at ryanyuan@canva.com. A lesser-known thread of his profile is his side projects/gitrepos (gocode.cc | backmir), suggesting ongoing engineering curiosity beyond formal publications.
12 years of coding experience