Zhihang Yuan
Papers
1
Total Citations
6
H-Index
1
About
Zhihang Yuan is a researcher specializing in efficient deep learning systems, with a particular focus on quantization, neural network compression, and 3D perception for autonomous driving. His work bridges the gap between high-performance AI models and their practical deployment on resource-constrained hardware, addressing one of the most pressing challenges in modern machine learning engineering. Yuan's most notable contribution, PillarHist, demonstrates his ability to innovate at the intersection of 3D object detection and quantization-aware training. By introducing a height-aware histogram-based pillar feature encoder, he tackled the fundamental challenge of deploying real-time LiDAR-based perception systems in autonomous vehicles and robotics platforms without sacrificing accuracy — a notoriously difficult balance to strike. This work reflects a broader commitment to making state-of-the-art models viable for onboard, edge-level deployment. Though early in its citation trajectory with 6 citations since 2025, PillarHist has already attracted the attention of researchers working at the frontier of autonomous systems. Yuan's research is particularly relevant to engineers and scientists seeking to deploy robust 3D detection pipelines under strict computational budgets, positioning him as an emerging voice in quantization-aware perception research.
Research Focus
Key Achievements
Top Papers
- 1