Naoto Soga

Papers

Total Citations

H-Index

About

Naoto Soga is a researcher at the forefront of embedded computer vision, specializing in efficient, real-time depth sensing for resource-constrained platforms. His primary research focuses on monocular depth estimation—a technique that infers 3D scene depth from a single camera image—and its hardware acceleration on FPGAs. Soga’s most cited work, "Fast Monocular Depth Estimation on an FPGA" (2020, 8 citations), addresses a critical bottleneck in robotics and autonomous systems: achieving reliable, pixel-wise depth perception without the cost or bulk of LiDAR. By optimizing deep learning models for low-power, small-area hardware, his contributions enable practical deployment in home robots, self-driving cars, and drones. This work underscores the growing demand for compact, affordable sensing solutions that maintain high accuracy. Soga’s research bridges the gap between algorithmic advances and real-world embedded applications, making 3D scene understanding more accessible. His achievements highlight a commitment to pushing the boundaries of on-device intelligence, where speed, efficiency, and minimal footprint are paramount. For students and engineers, Soga’s work exemplifies how to translate complex vision tasks into deployable, hardware-friendly systems.