Structure is one of the important features for image analysis and synthesis. In this talk, we will introduce three of our recent advances and applications with structure-aware analysis and compositing methods. In the first work, a novel foreground extraction method based on structure prediction and consistency check is presented. It tackles the problem of related methods for the structural scenes. The second work discusses how we enhanced the realism of sketch and portrait paintings through optimized image compositing, based only on a moderately sized photo collection. The third work presents an interesting tangible interface based on markerless tracking. A user can assemble real objects by intuitively following the live-feedback 3D AR instructions.
I-Chen Lin is an associate professor in the College of Computer Science, National Chiao Tung University (NCTU), Taiwan. He is currently a visiting scholar in Department of Computer Science, University of California, Davis, US.
Dr. Lin received the B.S. and Ph.D. degrees in computer science from National Taiwan University in 1998 and 2003, respectively. He joined NCTU as a faculty member in 2005. From August 2015 to July 2017, he also served as the Director of Industry Cooperation Division, Office of International Affairs. His research interests include computer graphics, computer vision, interactive techniques, and virtual/augmented reality.