Li, who spent more than 10 years at Microsoft Research in positions including head of Microsoft AI Asia's generative AI group, recently returned to China and has joined the newly established Institute of AI for Engineering at Tongji.
According to public records, Li received his bachelor's degree in computer science from Zhejiang University before earning a master's and PhD from Columbia University, where he was part of a five-person team that won the Grand Challenge 1st Place Award at ACM Multimedia 2013 in Barcelona.
The team, led by Brendan Jou and including Joseph G. Ellis, Daniel Morozoff-Abegauz and Columbia professor Shih-Fu Chang, was recognized for a paper on structured exploration of heterogeneous multimedia news sources.
After his PhD, Li joined Microsoft Research's U.S. headquarters in Redmond, Washington, where he held positions including principal researcher, principal architect and principal applied science manager, the South China Morning Post reported. His research focused on machine intelligence, multimodal content analysis and cloud computing.
He was one of seven co-authors of "Rethinking Classification and Localization for Object Detection," a paper presented at the IEEE/CVF Conference on Computer Vision and Pattern Recognition in June 2020 and led by Northeastern University's Yue Wu. The paper proposed a "double-head" neural network architecture for separating the distinct requirements of object classification and localization, and has become a widely cited reference in computer vision.
