ISD Seminar - Benchmarking Photometric Stereo Using a Broader Range of Real-world Data

[ISD Seminar - Benchmarking Photometric Stereo Using a Broader Range of Real-world Data] was presented by Prof. Boxin SHI, a Boya Young Fellow Associate Professor and Research Professor at Peking University on 9th April 2024. 

Since the first real-world benchmark dataset for photometric stereo (DiLiGenT) has been proposed in CVPR’16/TPAMI’19, almost all data-driven photometric stereo methods published after that evaluated their performances on it. But DiLiGenT’s scale and diversity can no longer meet the requirement of comprehensively evaluating newly proposed methods continuing to spring up, especially when recent models trained with large-scale synthetic data demonstrating strong generalization capabilities. If there is a sufficient amount of synthetic data and the model's capabilities are strong enough, do we still need to spend time and efforts in collecting real-world datasets? This talk will introduce recently released real-world photometric stereo benchmarks covering a broader range of objects, including 1) the DiLiGenT10^2  dataset containing objects of controlled  shape and reflectance variation through mechanical manufacturing (CVPR'22), 2) the DiLiGenT-Pi dataset containing planar objects with rich details through micron-level 3D scanning (ICCV'23), and 3) the DiLiGenRT dataset containing quantified reflectances by manufacturing hemispheres with varying degrees of roughness and translucency (CVPR'24). Based on the new series of benchmark datasets, state-of-the-art photometric stereo methods are systematically evaluated, and some open problems are discovered.

Boxin Shi is currently a Boya Young Fellow Associate Professor (with tenure) and Research Professor at Peking University, where he leads the Camera Intelligence Lab. He has also been a Young Scientist at Beijing Academy of Artificial Intelligence. He received the PhD degree from the University of Tokyo in 2013. From 2013 to 2017, he did research at MIT Media Lab, Singapore University of Technology and Design, Nanyang Technological University, and National Institute of Advanced Industrial Science and Technology. His research interests are computational photography and computer vision. He has published more than 200 papers, including 23 papers in TPAMI and 82 papers in CVPR/ICCV/ECCV.  His papers were awarded as Best Paper - Runner Up at ICCP 2015 and selected as Best Papers from ICCV 2015 for IJCV Special Issue. He received the Okawa Foundation Research Grant in 2021. He has served as an associate editor of TAPMI/IJCV and an area chair of CVPR/ICCV/ECCV. He is a Senior Member of the IEEE/CCF/CSIG, and a Distinguished Lecturer of APSIPA. Please access his lab’s website for more information: http://camera.pku.edu.cn