Xiaoyu Feng

Tsinghua University

Papers

1

Total Citations

4

H-Index

1

About

Xiaoyu Feng is at the forefront of energy-efficient deep learning hardware, with a specialized focus on multi-sensor perception for autonomous systems. His key research areas span scalable processor architectures, vision-point cloud fusion, and real-time AI acceleration for edge applications. Feng's most notable contribution is the development of a 28nm 1.2GHz scalable vision/point cloud deep fusion processor, which achieves an impressive 5.27TOPS/W efficiency. This work introduces a CAM-based universal mapping unit specifically designed for Bird's Eye View (BEVFusion) applications—a technique now setting benchmarks in autonomous driving and robot navigation. By addressing the substantial computational challenges of fusing camera and LiDAR data, his processor enables more accurate environmental perception while maintaining low power consumption. Though his 2024 paper has already garnered early citations, Feng's impact lies in pioneering hardware solutions that make complex multi-sensor AI practical for real-world deployment. His research bridges the critical gap between algorithmic advances in BEV fusion and the stringent energy constraints of autonomous platforms, positioning him as a rising leader in efficient deep learning acceleration.

Research Focus

Key Achievements

1
H-Index
1
Papers
4
Total Citations
4
Avg Citations/Paper
🏆 Most Cited Paper
A 28nm 1.2GHz 5.27TOPS/W Scalable Vision/Point Cloud Deep Fusion Processor with CAM-based Universal Mapping Unit for BEVFusion Applications
4 citations · 2024
📈 Most Prolific Year: 2024 (1 Papers)
🤝 Key Collaborators: 5
🏛 Institutions: Tsinghua University

Top Papers

  1. 1

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago