DocumentCode :
254220
Title :
The Synthesizability of Texture Examples
Author :
Dengxin Dai ; Riemenschneider, Hayko ; Van Gool, Luc
Author_Institution :
Comput. Vision Lab., ETH Zurich, Zurich, Switzerland
fYear :
2014
fDate :
23-28 June 2014
Firstpage :
3027
Lastpage :
3034
Abstract :
Example-based texture synthesis (ETS) has been widely used to generate high quality textures of desired sizes from a small example. However, not all textures are equally well reproducible that way. We predict how synthesizable a particular texture is by ETS. We introduce a dataset (21, 302 textures) of which all images have been annotated in terms of their synthesizability. We design a set of texture features, such as ´textureness´, homogeneity, repetitiveness, and irregularity, and train a predictor using these features on the data collection. This work is the first attempt to quantify this image property, and we find that texture synthesizability can be learned and predicted. We use this insight to trim images to parts that are more synthesizable. Also we suggest which texture synthesis method is best suited to synthesise a given texture. Our approach can be seen as ´winner-uses-all´: picking one method among several alternatives, ending up with an overall superior ETS method. Such strategy could also be considered for other vision tasks: rather than building an even stronger method, choose from existing methods based on some simple preprocessing.
Keywords :
computer vision; image texture; ETS method; data collection; example-based texture synthesis method; high quality texture generation; image property; winner-uses-all approach; Computer vision; Data collection; Learning systems; Pattern recognition; Robustness; Surface texture; Visualization; Homogeneity; Irregularity; Repetitiveness; Synthesizability; Texture Analysis; Texture Synthesis; Textureness;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Vision and Pattern Recognition (CVPR), 2014 IEEE Conference on
Conference_Location :
Columbus, OH
Type :
conf
DOI :
10.1109/CVPR.2014.387
Filename :
6909783
Link To Document :
بازگشت