Photographs & Drawings

Full Conference  One-Day Full Conference

Friday, 12 December
10:30 - 12:00
Room 302
Session Chair: Diego Gutierrez

Forward Lean - Deriving Motion Illustrations From Video
Forward Lean extracts moving objects from a video sequence and then illustrates the objects' motions in a single static image by shearing the image into the direction of its motion.

Marc Nienhaus
mental images GmbH

Holger Winnemöller
Adobe Research Inc.

Bruce Gooch
University of Victoria
Automatic 3D Caricature Generation By Learning in Enlarged Manifold Space
Lack of samples makes it challenging to generate 3D caricatures by machine learning. In this method, a training set is enlarged by reconstructing 3D caricatures, and then a regressive model is learnt by manifold regularisation.

Junfa Liu
Chen Yiqiang
Chinese Academy of Sciences

Chunyan Miao
Nanyang Technological University

Wen Gao
Peking University
Visualising Adaptive Clusters of Digital Photographs
With this visualisation method, which considers semantic flow in each cluster, users can select references from a specific camera for smart visualisation if concurrent photos are used as input data.

Chuljin Jang
Hwan-Gue Cho
Pusan National University
Clean up Your Image Using Internet Photo Collections
An algorithm that uses images from internet photo collections to remove user-identified occlusions in an image and faithfully reconstruct the image data that should have been displayed.

Hanieh Taipalus
Helsinki University of Technology

Satoshi Kondo
Matsushita Electric Industrial Co., Ltd.

Takafumi Aoki
Tohoku University