Lucas de Vries
Papers
1
Total Citations
20
H-Index
1
About
Lucas de Vries is a leading researcher in efficient deep learning and real-time computer vision, whose work addresses the critical challenge of deploying high-performance neural networks in resource-constrained environments. His seminal 2019 paper, "Evolving Efficient Deep Neural Networks for Real-time Object Recognition," has garnered 20 citations and introduced a groundbreaking approach to accelerating DNNs by systematically reducing their parameter counts without sacrificing accuracy. This work directly tackles the computational bottleneck of deep networks with millions or billions of parameters, enabling real-time object recognition on edge devices. De Vries’s contributions lie at the intersection of neural architecture search, model compression, and evolutionary optimization, where he has pioneered methods that automatically discover compact, efficient network designs. His research has significant implications for autonomous systems, mobile robotics, and embedded AI, making deep learning practical for latency-sensitive applications. By demonstrating that evolutionary algorithms can effectively prune and refine deep architectures, de Vries has opened new pathways for sustainable, high-speed AI deployment, establishing himself as a key innovator in the push toward efficient, real-world neural computation.
Research Focus
Key Achievements
Top Papers
- 1Evolving Efficient Deep Neural Networks for Real-time Object Recognition20 citations · 2019