| Real-time Integration of Speech, Gesture, Graphics and Data-base (1995) | |||||||||||||
Abstract | |||||||||||||
| Introduction A real-time system is proposed for integrating spotting recognition results of human gestures from motion images and spontaneous speech. The integration of the two kinds of spotting is carried out through the computation of an automaton which realizes a new architecture called " Grainfree and parallel architecture." 2. Real-time Recognition System Based on Grain-free Parallel Architecture word sentence concept phrase intention Input: (Speech, Gesture) Output: (Speech, Symbol, Graphic@Image) Figure 1: Grain-free and parallel architecture for human thinking. We propose a new architecture called "Grainfree Parallel architecture" which is strongly coupled with the capability of real-time response for a man-machine interface system operated by spotting recognition of speech and gesture. The requirement for the system leads to a new architecture (shown in figure 1) which is analogous to the architecture proposed by Brooks [1] in robotics. T | |||||||||||||
Publication details | |||||||||||||
| |||||||||||||