Papers
4
Total Citations
1,275
H-Index
3
About
Ningyi Xu is a leading researcher at the intersection of efficient deep learning hardware and embodied AI, whose work has fundamentally shaped how neural networks are deployed on resource-constrained platforms. His seminal 2016 paper, “Going Deeper with Embedded FPGA Platform for Convolutional Neural Network,” has amassed over 1,260 citations and established a foundational framework for accelerating CNNs on FPGAs, enabling high-performance computer vision on edge devices. This work remains a cornerstone reference for hardware-software co-design in embedded AI. More recently, Xu has pioneered advances in 3D perception and robotics. His work on SpOctA introduces an octree-encoding-based accelerator for 3D sparse convolution networks, dramatically improving efficiency for point-cloud processing in autonomous driving and AR/VR. Demonstrating a forward-looking vision, Xu’s 2025 paper on SARO explores the integration of vision-language models with quadruped robot navigation, enabling space-aware terrain crossing in 3D environments. Through a career that bridges hardware acceleration, sparse computation, and foundation-model-driven robotics, Ningyi Xu continues to drive innovation in efficient, intelligent systems for real-world deployment.
Research Focus
Key Achievements
Top Papers
- 1Going Deeper with Embedded FPGA Platform for Convolutional Neural Network1,260 citations · 2016
- 2
- 3
- 4