Modelling the Creation of Verbal Indoor Route Descriptions for Visually Impaired Travellers

Abstract

Within the field of computer-supported indoor navigation for visually impaired people, the generation of effective verbal route descriptions and directions to be given to the visually impaired person remains challenging. This paper provides a formal and innovative model for the creation of indoor verbal route descriptions (VRDs) with the iterative methodologies from user-centred design, focused on those with visual impairment (VI), and emphasising sufficient evaluation. As one step towards fully automated generation of verbal directions, four sighted persons were tasked with generating VRDs for two routes at Saarbrücken University. The generated VRDs were evaluated with feedback by 11 VIs. This was used to improve the VRDs which were evaluated again by 3 VIs, with the set of 7 directions having the overall result rated as medium. Implications and main pitfalls of the current model are discussed.