作者: H.-J. Boehme , U.-D. Braumann , A. Brakensiek , A. Corradini , M. Krabbes
关键词:
摘要: Recently there is an increasing interest in video based interface techniques, allowing more natural interaction between users and systems than common devices do. We present a neural architecture for user localisation, embedded within complex system visually-based human machine (HMI). User localisation absolute prerequisite to HMI. Due the main objective, greatest possible robustness of as well whole visual under highly varying environmental conditions, we propose multiple cue approach. This approach combines features facial structure, head shoulder contour, skin color motion, with multiscale representation. The selection image region most likely containing then realised via WTA-process Preliminary results show reliability