Audiowiz: nearly real-time audio transcriptions

  • Authors:
  • Samuel White

  • Affiliations:
  • University of Rochester, Rochester, NY, USA

  • Venue:
  • Proceedings of the 12th international ACM SIGACCESS conference on Computers and accessibility
  • Year:
  • 2010

Quantified Score

Hi-index 0.01

Visualization

Abstract

Existing automated transcription solutions filter out environmental noises and focus only on transcribing the spoken word. This leaves deaf and hard of hearing users with no way of learning about events that provide no spoken information such as the sounds produced by a faulty appliance or the barked alert of a dutiful guard dog. In this paper we present AudioWiz, a mobile application that provides highly detailed audio transcriptions of both the spoken word and the accompanying environmental sounds. This approach is made possible by harnessing humans to provide audio transcriptions instead of more traditional automated means. Web-workers are recruited automatically in nearly real-time as dictated by demand.