Meixuan Lu
Papers
1
Total Citations
4
H-Index
1
About
Meixuan Lu’s research lies at the intersection of computer architecture and intelligent vision systems, with a focus on accelerating stereo vision processing for real-world applications like robotics, autonomous vehicles, and augmented reality. Her most-cited work, “Dadu-SV: Accelerate Stereo Vision Processing on NPU” (2022), addresses a critical bottleneck in modern vision pipelines: the heavy computational demands of both classic semiglobal matching (SGM) and deep convolutional neural networks (CNNs). By designing a specialized acceleration framework for neural processing units (NPUs), Lu demonstrates how to achieve high-performance binocular vision without sacrificing efficiency—a key step toward making advanced perception systems practical for edge devices. With 4 citations and growing recognition, this work showcases her ability to bridge algorithmic complexity and hardware constraints. Lu’s contributions are particularly valuable for students and engineers seeking to understand how to optimize computer vision workloads on emerging neural accelerators, blending theoretical insight with hands-on architectural innovation. Her research continues to influence the development of faster, more energy-efficient vision systems for autonomous navigation and immersive AR experiences.
Research Focus
Key Achievements
Top Papers
- 1Dadu-SV: Accelerate Stereo Vision Processing on NPU4 citations · 2022