Binh‐Son Hua

Singapore University of Technology and Design, The University of Tokyo

Papers

Total Citations

434

H-Index

About

Binh-Son Hua is a leading researcher in 3D computer vision and robotics, with a focus on real-time semantic scene understanding. His major contributions center on developing efficient methods for dense 3D reconstruction and semantic segmentation of indoor environments, enabling autonomous systems like drones and assistant robots to perceive and navigate complex spaces in real time. Hua’s most impactful work includes the creation of **SceneNN** (346 citations), a richly annotated RGB-D dataset that addressed the critical lack of fine-grained ground truth in existing benchmarks, providing a vital resource for training and evaluating scene understanding models. He also pioneered **real-time progressive 3D semantic segmentation** techniques (79 citations), which allow for on-the-fly, high-quality labeling of indoor scenes during reconstruction—a breakthrough for practical robotics applications. By combining robust algorithmic design with high-quality data, Hua has significantly advanced the field’s ability to achieve both speed and accuracy in 3D perception. His work is essential reading for researchers developing autonomous systems that must interpret and interact with dynamic indoor environments.