Vouch-T: Multimodal text input for mobile devices using voice and touch

Abstract

Entering text on a small mobile device is not easy due to the relatively small screen space and the fat finger problem. We consider a multimodal text input method, called Vouch-T (Voice + tOUCH - Text) combining the touch and voice input in complementary way. With Vouch-T, the user makes an approximate touch among densely distributed alphabetic keys, accompanied with the voice input to effectively disambiguate the target among possible candidates. We have assessed the potential of Vouch-T in terms of the usability and performance compared to the conventional touch-only based method. We considered two types of text input layout, namely, the QWERTY and 3 × 4 telephone keypad on two sizes of mobile devices, the smart phone and smart watch. The comparative simulation experiment validated that the multimodal approach of Vouch-T improves the input performance and usability for the smaller-sized smart watch, but only marginally for the larger smart phone.