A trajectory-based approach for device independent gesture recognition in multimodal user interfaces

With the rise of technology in all areas of life new interaction techniques are required. With gestures and voice being the most natural ways to interact, it is a goal to also support this in human-computer interaction. In this paper, we introduce our approach to multimodal interaction in smart home environments and illustrate how device independent gesture recognition can be of great support in this area. We describe a trajectory-based approach that is applied to support device independent dynamic hand gesture recognition from vision systems, accelerometers or pen devices. The recorded data from the different devices is transformed to a common basis (2D-space) and the feature extraction and recognition is done on this basis. In a comprehensive case study we show the feasibility of the recognition and the integration with a multimodal and adaptive home operating system.