MAS 2013 Proceeding

Using conversive hidden non-markovian models for multi-touch gesture recognition

Authors:   Tim Dittmar, Claudia Krull, Graham Horton


With the current boom of multi-touch devices the recognition of multi-touch gestures is becoming an important field of research. Performing such gestures can be seen as a stochastic process, as there can be many little differences between each execution. Therefore stochastic models like Hidden Markov Models have been already utilized for gesture recognition. Although the modelling possibilities of Hidden Markov Models are limited, they achieve an acceptable recognition quality. But they have never been tested with gestures that only differ in execution speed. Therefore we propose to use Conversive Hidden non-Markovian Models for multi-touch gesture recognition. This extension of Hidden Markov Models enhances the modelling possibilities and adds timing features. In this work two multi-touch gesture recognition systems were developed and implemented based on these two model types. Experiments with a set of similar gestures show that the proposed model is a good and competitive alternative and can even be better than Hidden Markov Models.

I3M  Scientific Sponsors

I3M  Industrial Sponsors

I3M  Media Sponsors