The 4th ACM SIGGRAPH Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia
Conference 12-15 December • Exhibition 13-15 December • Hong Kong Convention & Exhibition Centre
 

Technical Papers

Image Mix and Match

Wednesday, 14 December 14:15 - 16:00 |  Convention Hall B

Session Chair
Shi Min Hu

Data-driven Visual Similarity for Cross-domain Image Matching - Picture

Data-driven Visual Similarity for Cross-domain Image Matching


Data-driven image-based methods require a similarity measure for finding similar images. We learn query-specific similarity measures where the importance of different parts of an image is based on our notion of "data-driven uniqueness''. Our approach enables matching across diverse visual domains involving images, sketches and paintings.


Abhinav Shrivastava, Carnegie Mellon University
Tomasz Malisiewicz, Carnegie Mellon University
Abhinav Gupta, Carnegie Mellon University
Alexei A. Efros, Carnegie Mellon University


Arcimboldo-like Collage Using Internet Images - Picture

Arcimboldo-like Collage Using Internet Images


Make yourself disguised with extravagant yet lucid form to show attractive
appearance. This is our fascinating Arcimboldo-like collage.


Hua Huang, Xi'an Jiaotong University
Lei Zhang, Xi'an Jiaotong University
Hong-Chao Zhang, Xi'an Jiaotong University


Semantic Colorization with Internet Images - Picture

Semantic Colorization with Internet Images


We proposed a novel colorization method that utilizes internet photos and image filtering to minimize user effort and facilitate accurate color transfer. Both image filtering and colorization results were shown to outperform related methods.


Alex Yong-Sang Chia, Agency for Science, Technology and Research
Shaojie Zhuo, National University of Singapore
Raj Kumar Gupta, Nanyang Technological University
Yu Wing Tai, Korea Advanced Institute of Science and Technology
Siu-Yeung Cho, Nanyang Technological University
Ping Tan, National University of Singapore
Stephen Lin, Microsoft Research Asia


Rendering Synthetic Objects into Legacy Photographs - Picture

Rendering Synthetic Objects into Legacy Photographs


We propose a method to realistically insert synthetic objects into existing photographs without requiring access to the scene or any additional scene information. Through a user study, we show that our results are often confused with real scenes, and preferred to other state-of-the-art methods.


Kevin Karsch, University of Illinois at Urbana-Champaign
Varsha Hedau, University of Illinois at Urbana-Champaign
David Forsyth, University of Illinois at Urbana-Champaign
Derek Hoiem, University of Illinois at Urbana-Champaign