Hyunsoo Park
Papers
1
Total Citations
9
H-Index
1
About
Hyunsoo Park is a rising researcher in computer vision and 3D scene understanding, with a focus on reconstructing complex hand-object interactions from minimal visual input. His most notable contribution, "HandNeRF," introduces a pioneering method to learn interaction priors that enable the reconstruction of a full 3D hand-object scene from a single RGB image—a task long considered daunting due to depth ambiguity and severe occlusions. This work, published in 2024 and already garnering 9 citations, demonstrates Park's ability to tackle fundamental challenges in neural rendering and human-centric modeling. By generating training data and inference pipelines for such constrained settings, Park's research bridges the gap between 2D observations and realistic 3D interaction reconstruction, with implications for augmented reality, robotics, and human-computer interaction. His approach not only advances the field of implicit neural representations but also offers practical solutions for real-world applications where multi-view data is unavailable. As an emerging scholar, Park's work signals a promising trajectory in synthesizing high-fidelity 3D content from sparse inputs, making him a researcher to watch in the evolving landscape of vision-based interaction modeling.
Research Focus
Key Achievements
Top Papers
- 1