Papers
2
Total Citations
6
H-Index
2
About
Zherong Liu is a researcher at the forefront of energy-efficient computer vision and embedded AI systems, with a focus on hardware-software co-design for real-time visual intelligence. His work centers on developing specialized accelerators that bridge the gap between algorithmic performance and hardware efficiency, particularly for resource-constrained platforms like mobile robots and UAVs. Liu’s most cited paper, “A High-Performance ORB Accelerator with Algorithm and Hardware Co-design for Visual Localization” (2024, 4 citations), introduces a novel approach to accelerating the classic ORB algorithm for visual localization—a critical component in AR/VR and autonomous navigation. By tightly integrating algorithm optimization with dedicated hardware, his design achieves significant improvements in throughput and energy efficiency. In his earlier work, “An Energy-Efficient Visual Object Tracking Processor Exploiting Domain-Specific Features” (2023, 2 citations), Liu demonstrates how leveraging domain-specific knowledge can outperform general AI accelerators for visual object tracking, a key technology in intelligent surveillance and robotics. His contributions are particularly notable for pushing the boundaries of on-device AI, enabling complex vision tasks to run efficiently on battery-powered systems. As the demand for edge intelligence grows, Liu’s research offers a compelling path toward practical, high-performance visual computing in the real world.
Research Focus
Key Achievements
Top Papers
- 1
- 2