Youki Sada

National Institute of Information and Communications Technology

Papers

Total Citations

H-Index

About

Youki Sada is a researcher at the forefront of embedded computer vision, specializing in efficient deep learning architectures for real-time 3D scene understanding. His primary contributions lie in monocular depth estimation—a critical technology for enabling autonomous systems like home robots, self-driving cars, and drones to perceive their environment using only a single camera. Sada’s most cited work, "Fast Monocular Depth Estimation on an FPGA" (2020), tackles the challenge of deploying complex neural networks on resource-constrained hardware. By optimizing algorithms for Field-Programmable Gate Arrays, he demonstrated that high-quality pixel-wise depth maps can be generated with low latency and minimal power consumption, making depth sensing practical for compact, embedded platforms. This achievement bridges the gap between cutting-edge computer vision and real-world deployment, addressing the need for reliable, low-cost perception in mobile robotics. With 8 citations, his work has already influenced subsequent research in hardware-accelerated vision systems. Sada’s research continues to push the boundaries of what is possible on edge devices, promising smarter, more autonomous machines that can navigate and interact with the world safely and efficiently.