Summary
Xiao Yang is a Principal Research Scientist with nine years of experience bridging academic computer vision research and large-scale product engineering across ByteDance, Canva, and Facebook. He holds a PhD from UNC Chapel Hill and built influential deep learning methods for fast image registration and medical image analysis under Prof. Marc Niethammer. At ByteDance he led AIGC media generation efforts—shipping 20+ high-impact visual effects, pretraining Seeddream models, and delivering accelerator models like SDXL-Lightning and Animatediff-Lightning that powered company products and topped Hugging Face rankings. Now at Canva CORE he focuses on image/video generative models for the design space while continuing to push for orders-of-magnitude efficiency gains in video generation. Known for turning cutting-edge research into production wins, Xiao blends numerical rigor, systems-level engineering, and product sensibility across vision and AR domains.
9 years of coding experience