Dun-Yu Hsiao

Papers

Total Citations

H-Index

About

Dun-Yu Hsiao is a rising researcher in 3D computer vision and robotics, whose work focuses on the critical challenge of efficient 3D scene representation. His most cited paper, "Learning View Selection for 3D Scenes" (2021, 8 citations), tackles the fundamental problem of optimal viewpoint sampling—a task essential for reconstructing and understanding 3D objects and environments. Rather than relying on brute-force dense view collections, Hsiao reformulates this as a learnable set cover problem, introducing a data-driven approach that intelligently selects the most informative views. This work bridges classical geometric optimization with modern deep learning, offering a more efficient pathway for 3D vision pipelines. While still early in his career, Hsiao's contributions are particularly relevant for applications in robotic perception, autonomous navigation, and augmented reality, where computational efficiency and accuracy are paramount. His research demonstrates a keen ability to identify and solve foundational bottlenecks in 3D understanding, positioning him as a promising voice in the field. As the demand for robust 3D scene analysis grows, Hsiao's innovative view selection framework provides a scalable solution that could influence future generations of 3D vision systems.