Naoya Chiba
Papers
3
Total Citations
15
H-Index
2
About
Naoya Chiba is a researcher pushing the boundaries of 3D computer vision and robotics, with a focus on solving the most challenging problems in scene understanding and measurement. His work is defined by a unique combination of physical simulation, advanced optics, and cutting-edge neural rendering. Chiba’s most impactful contribution, “Ultra-Fast Multi-Scale Shape Estimation of Light Transport Matrix for Complex Light Reflection Objects” (11 citations), tackles the notoriously difficult task of measuring objects with specular reflections or subsurface scattering—surfaces that break traditional 3D scanning methods. By modeling complex light paths, he enables accurate shape estimation for these “complex light reflection objects.” More recently, with “NeuralLabeling” (2024), Chiba has developed a versatile toolset that leverages Neural Radiance Fields (NeRFs) to automate the labeling of 3D scenes, generating a comprehensive suite of annotations from segmentation masks to 6DOF poses. This work promises to dramatically accelerate the creation of high-quality training data for vision systems. His foundational work on generating point clouds for bin-picking scenes using physical simulation (2017) further underscores his commitment to bridging simulation and reality for industrial robotics. Chiba’s research is essential reading for anyone working in 3D reconstruction, robotic manipulation, or synthetic data generation.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3