Chuxi Li

South China University of Technology, Northwestern Polytechnical University

Papers

Total Citations

H-Index

About

Chuxi Li is a researcher whose work bridges the critical fields of multi-robot systems and efficient deep neural network (DNN) hardware acceleration. In robotics, Li has made foundational contributions to cooperative localization, developing Extended Kalman Filter (EKF)-based algorithms that enable teams of robots—composed of a leader and followers—to accurately determine their positions using only local sensors and neighbor-to-neighbor communication. This distributed approach, detailed in papers from 2016, addresses a core challenge in autonomous navigation for applications like search-and-rescue and intelligent transportation. More recently, Li has ventured into the hardware-software co-design of DNN accelerators, proposing a novel "memory-computing decoupling" architecture. This design, published in 2022, allows edge devices to efficiently run multiple DNNs simultaneously by adaptively arranging data to match each subtask's preferred dataflow—a critical advancement for resource-constrained platforms like autonomous vehicles and intelligent robots. With over 16 total citations across these key works, Li’s research demonstrates a consistent focus on enabling robust, real-time intelligence for distributed, multi-agent systems operating under real-world constraints.