The Topo-Speech Algorithm: An intuitive Sensory Substitution for Spatial Information

Abstract

Is it possible to quickly and reliably understand the position of objects in space without vision? This is one of the biggest challenge blind people face daily. We developed a novel algorithm called the Topo-Speech which conveys the spatial position of objects via speech manipulations. We ran a pilot study on blindfolded sighted adults (n=5) to test the extent to which users can locate objects’ spatial positions after short training with the Topo-Speech, as well as their ability to locate untrained spatial positions. Participants were trained for ~30 minutes on the detection of objects’ positions on a 3×3 grid. Then they were tested on the same spatial locations (though using different stimuli). Finally, participants were tested on identifying the positions of objects on a 5×5 grid (i.e. additional spatial locations) without any specific training. Our results showed that participants performed significantly above chance for both trained and untrained spatial positions. This in turn suggests the feasibility of the Topo-Speech to convey spatial related information via a non-visual channel and prompt to quickly test such approach with people who are visually impaired.