Title :
Evolutionary algorithms for gesture segmentation
Author :
Handa, Hisashi ; Kawakam, Hiroshi
Author_Institution :
Grad. Sch. of Natural Sci. & Technol., Okayama Univ., Okayama, Japan
Abstract :
From the viewpoint of “UBEN-EKI,” we have developed an inconvenient input method for mobile devices, which provides us with chances of exploration. In the proposed mobile device, users provide a large number of motions to systems, which should be learnt. In this paper, we constitute evolutionary algorithms to automatically extract frequently-observed motions from whole the motion sequence.
Keywords :
evolutionary computation; feature extraction; gesture recognition; image motion analysis; image segmentation; image sequences; mobile computing; time warp simulation; FUBEN-EKI; dynamic time warping; evolutionary algorithm; frequently-observed motion extraction; gesture segmentation; mobile device; motion sequence; Data mining; Evolutionary computation; Humans; Mobile handsets; Speech; Synthesizers; Three dimensional displays; Dynamic Time Warping; Evolutionary Algorithms; Frequently-Observed Motion Extraction;
Conference_Titel :
SICE Annual Conference 2010, Proceedings of
Conference_Location :
Taipei
Print_ISBN :
978-1-4244-7642-8