Tuning Autoregressive Texture Stimuli on an Electrostatic Friction Display to Enhance Perceived Realism

Electrostatic friction displays present tactile textures on touch panels by controlling surface friction. For presenting fabric textures, data-driven models are commonly used to simulate friction force fluctuations generated during finger rubbing. In this study, we investigate whether tactile realism can be enhanced in a tailor-made manner by adding friction modulation components in low-, mid-, and high-frequency bands to texture stimuli generated by a data-driven autoregressive model. Using an electrostatic friction-based tactile display, we conducted a user study in which participants rated the perceived realism of tactile stimuli corresponding to two fabric types: cotton and silk. For cotton, the stimulus with additional friction modulation was rated as more realistic than those generated by the autoregressive model alone. In contrast, no significant difference in perceived realism was observed for silk. These results suggest that tactile realism can be improved by expert-guided tuning of data-driven tactile stimuli, depending on the material characteristics.