Papers
1
Total Citations
35
H-Index
1
About
Tong Geng is a leading researcher in efficient deep learning inference, specializing in binarized neural networks (BNNs) and hardware acceleration. His most cited work, "LP-BNN: Ultra-low-Latency BNN Inference with Layer Parallelism" (2019, 35 citations), addresses a critical bottleneck in deploying DNNs for real-time applications like autonomous driving and robotic control. Geng’s key contribution is the development of a novel layer-parallel architecture that dramatically reduces inference latency in BNNs, enabling ultra-low-latency performance without sacrificing accuracy. This work has been foundational for researchers and engineers seeking to push the boundaries of approximate computing and edge AI. By tackling the trade-off between precision and speed, Geng has helped pave the way for practical, high-performance neural network deployment in latency-sensitive domains. His research continues to influence the design of efficient, hardware-aware deep learning systems, making him a notable figure in the intersection of computer architecture and machine learning.
Research Focus
Key Achievements
Top Papers
- 1LP-BNN: Ultra-low-Latency BNN Inference with Layer Parallelism35 citations · 2019