Kazuki Fukitani

Papers

Total Citations

H-Index

About

Kazuki Fukitani is a researcher focused on advancing 3D object detection for robotics and autonomous systems, with particular expertise in point cloud-based deep learning. His most-cited work, "3D Object Detection Using Improved PointRCNN" (2022), addresses critical limitations in two-dimensional detection methods, which lack the depth information essential for indoor robot navigation and grasping tasks. By enhancing the PointRCNN architecture, Fukitani improved the accuracy and efficiency of detecting objects in three-dimensional space from LiDAR or depth sensor data. Though his citation count is still growing—with 4 citations on his top paper—his contributions are timely, tackling real-world challenges in robotics where precise spatial understanding is paramount. His research bridges the gap between 2D detection applications (like surveillance and medical imaging) and the 3D perception needs of autonomous systems. Fukitani’s work is particularly notable for targeting practical deployment in indoor environments, where robots must navigate cluttered spaces and interact with objects. As 3D vision becomes increasingly vital for embodied AI, his improvements to PointRCNN represent a foundational step toward more reliable robotic perception.