Papers
5
Total Citations
29
H-Index
4
About
Tsuyoshi Isshiki is a researcher at the forefront of robotics, embedded systems, and efficient machine learning hardware. His work spans the critical intersection of autonomous navigation and power-efficient deep learning acceleration. In visual SLAM, Isshiki advanced relocalization speed and accuracy by integrating object detection, a contribution that has garnered 13 citations and addresses a key bottleneck for indoor mobile robots. He has also developed practical underwater positioning systems for ROVs using trilateration, demonstrating his versatility in real-world robotic applications. A standout achievement is his end-to-end implementation of YOLOv8 on a RISC-V architecture, which achieved power-efficient, runtime-configurable object detection—a significant step for edge AI. Additionally, he has contributed to ROS-based mobile robot pose planning for autonomous 3D reconstruction and designed scalable FPGA hardware for gradient boosted tree training, targeting real-time, low-power machine learning. Isshiki’s work is characterized by a hands-on, systems-level approach, bridging algorithmic innovation with efficient hardware implementation, making his research highly relevant for students and engineers working on autonomous systems and embedded AI.
Research Focus
Key Achievements
Top Papers
- 1Improving Relocalization in Visual SLAM by using Object Detection13 citations · 2022
- 2
- 3A Power-efficient end-to-end Implementation of YOLOv8 Based on RISC-V4 citations · 2023
- 4
- 5