Study on gesture recognition system using posture classifier and Jordan recurrent neural network
Hiroomi Hikawa, Yusuke Araga · 2011
This paper proposes a Jordan recurrent neural network (JRNN) based dynamic hand gesture recognition system. A set of allowed gestures is modeled by a sequence of representative static images, i.e., postures. The proposed system first classifies the input postures contained in the input video frames, and the resulting posture indexes are fed to the JRNN that can detect dynamic temporal behavior. The feasibility of the proposed system and its characteristics are examined by experiments. Especially the effects of the posture classification performance and the gesture speed are studied. Experimental results show that the system recognize 10 gestures with the accuracy of 95%.