Zhiheng Yue
Papers
2
Total Citations
13
H-Index
2
About
Zhiheng Yue is a rising researcher at the forefront of energy-efficient hardware design for computer vision, with a focused expertise in **computation-in-memory (CIM) architectures** and **stereo vision processing**. His primary contribution lies in revolutionizing the **cost-volume construction** kernel—a critical but computationally expensive step in depth perception used in robotics, autonomous driving, and AR/VR. Yue introduced the **XOR-derived similarity-aware CIM**, a novel hybrid-domain approach that replaces traditional, power-hungry arithmetic with bitwise operations, dramatically reducing the energy and data movement overheads associated with large parameter sizes. His foundational work, "7.7 CV-CIM," published in 2023, has already garnered **10 citations**, demonstrating immediate impact in the VLSI and circuit design community. A subsequent 2024 paper extends this concept, solidifying his role in advancing efficient, real-time 3D reconstruction. By tackling the bottleneck of continuous memory access in stereo matching, Yue’s innovations enable high-performance vision systems to operate within strict power budgets, making him a key contributor to the next generation of intelligent edge devices.
Research Focus
Key Achievements
Top Papers
- 1
- 2