Research projects initiated by academics from universities or research institutions, and sponsored by Microsoft Research Asia in Singapore, serve to foster partnerships and collaboration with researchers pursuing technological advances in computer science.
Symmetric Architecture Modeling with a Single Image
Assistant Professor Ping Tan, National University of Singapore
The preservation of historical architectures is vital in support of the study and understanding of human civilization and culture by generations to come. However, the preservation and restoration of these architectures often face a dual-sided challenge—how to maintain the original architecture while repairing deterioration. The use of digital technologies to archive and restore historical architectures provides a solution to this dilemma. Ping Tan, assistant professor, National University of Singapore, collaborated with researchers at Microsoft Research Asia and proposed a new methodology to provide a solution to this dilemma by restoring historical architectures with three-dimensional (3-D) texture-mapped models. With this method, Tan was able to reconstruct traditional Chinese architecture buildings such as the Pavilion of Manifest Benevolence in the Forbidden City, Beijing, from a single input image. Read more...
This work has been published by Nianjuan Jiang, Ping Tan, and Loong Fah Cheong in “Symmetric Architecture Modeling with a Single Image,” ACM Transaction on Graphics (TOG) and Proc., SIGGRAPH Asia 2009.
The video illustrates the camera’s calibration from a single image followed by the reconstruction of a set of 3-D points according to the calibration and architecture symmetry. Next, the user interactively maps out structural components, such as roofs and walls, to build an initial 3-D model. The user can also add more geometric detail, such as roof tiles and handrails, or insert predefined primitives, such as pillars and staircases. Finally, the recovered model is textured according to the input image. Texture synthesis is used to improve texture quality at the foreshortened and occluded regions.
Semantic Colorization with Internet Images
Dr. Alex Yong-Sang Chia, Institute for Infocomm Research, A*STAR, Singapore
Alex Yong-Sang Chia, principal investigator at the Institute for Infocomm Research (I2R), in collaboration with Microsoft Research Asia, created a colorization system that takes advantage of the rich image content from the Internet. Chia investigated colorizing gray-scale photographs by focusing on low-level image restoration such as image color correction, denoising and deblurring, and high-level image editing such as image style transfer and image composition. This resulted in natural looking color images and demonstrated higher quality and effectiveness in comparison with other techniques.
This work has been published by Alex Yong-Sang Chia, Shaojie Zhuo, Raj Kumar Gupta, Yu Wing, Shih-Yeung Cho, Ping Tan, and Steve Lin in “Semantic Colorization with Internet Images,” SIGGRAPH Asia 2011.
The video presents a system for colorizing an input gray-scale image by using color photos downloaded from the Internet.