The 2023 Asiagraphics (AG) Young Researcher Award was presented to Dr. Peng-Shuai Wang from Peking University, China. The winner of this award was selected by the award jury chaired by Prof. Ming Lin (UMD College Park) and Prof. Leif Kobbelt (RWTH Aachen).
Dr. Peng-Shuai Wang is a tenure-track Assistant Professor at Peking University. Before joining Peking University in 2022, he was a senior researcher in Microsoft Research Asia. He got Ph.D. degree from the Institute for Advanced Study at Tsinghua University in 2018, under the supervision of Dr. Baining Guo. Dr. Wang has done a series of remarkable research works on fundamental network structures and algorithms for 3D shape analysis and generation, which significantly advance the state-of-the-art of 3D geometric deep learning and make impactful contributions to both computer graphics and computer vision.
Dr. Wang’s research on Octree-based Sparse Convolutional Networks (O-CNN, SIGGRAPH 2017) lays a solid foundation for learning-based 3D shape analysis and generation and attracts considerable attention in the research field. O-CNN significantly reduces the computational and memory complexity of 3D deep learning from O(N^3) to O(N^2) and has been widely used in various 3D learning tasks, including 3D classification, segmentation, and detection. His work on Adaptive O-CNN (SIGGRAPH Asia 2018) also greatly improves the state-of-the-art for shape representation and generation.
To generate continuous surfaces and further improve the reconstruction of geometric details, Dr. Wang proposed Dual Octree Graph Networks (SIGGRAPH 2022) that offers an adaptive deep representation of 3D volumetric fields and associated graph neural networks, which greatly improves the efficiency and performance for shape generation and reconstruction. As transformer-based backbone networks have been widely used in 2D vision and NLP fields, Dr. Wang recently proposed OctFormer (SIGGRAPH 2023) that is not only significantly faster than previous point cloud transformers, but also achieves state-of-the-art performances in various 3D understanding tasks.
Additionally, Dr. Wang is also well known by his outstanding works on traditional and learning-based digital geometry processing, including his early work on learning-based mesh denosing (SIGGRAPH Asia 2016), and interactive geometric feature editing (SIGGRAPH Asia 2015), as well as his recent works on geodesic distance computation with graph neural networks (GeGNN in SIGGRAPH Asia 2023).
Dr. Wang also actively serves the graphics communities as the PC members of graphics conferences (e.g. Eurographics 2024, CVM 2023 & 2024.), and the paper reviewers of graphics and vision conferences and journals, such as ACM SIGGRAPH/TOG, IEEE TVCG, CVPR and CVMJ.
