Can attention decoding help those who need it the most?

Barbara Shinn-Cunnigham, Carnegie Mellon University

Abstract

Decoding what source a listener wants to attend in a crowded setting could revolutionize next-generation assistive listening devices — and rapid advances in the efficacy of decoding algorithms make it seem like such aids will become commercially available in the near future. However, the very listeners who most need such aid (e.g., those with hearing impairment, using cochlear implants, suffering from deficits in attentional control, or even coping with normal aging) may not produce clear markers of attentional focus. How can one decode attention in a listener for whom attentional focus is poor?  This presentation will raise questions we need to consider in order to translate the promise of the “smart hearing aid of the future” into a robust device that can aid the heterogenous mixture of individuals who have difficulty communicating in crowded settings.