VoiceMath project

Videos, as objects, and in partcular video lectures (frequently used in the pandemic period) have several accessibility problems: adequate subtitles are needed by people with hearing impairment, and transcriptions by people with motor or visual impairment, or specific learning disorders, in order for them to take notes. There are currently several systems for automatic transcription and subtitling, but none of these is able to manage the presence of formulas in a correct way: for scientific content videos to be accessible, the formulas need to be transcribed in a specific language (LaTeX or MathML ).

VoiceMath project, founded by Fondazione CRT, Dipartimento di Matematica “G. Peano” and Direzione SIPE (UniTO), aims to develop a software for textual and LaTeX transcription of university video lectures containing formulas, useful for people with motor, hearing and visual impairment and to people with specific learning disorders. This software has been implemented in collaboration with H-Farm Innovation, and is currently patent pending at University of Turin.

  • Tiziana Armano, Gabriele Baratto, Cristian Bernareggi, Marina Bosco, Manuela Caramagna, Anna Capietto, Mattia Ducci, Alessandro Mazzei, Antonio Mazzei, Francesco Tarasconi, Matematica a voce parte 1: trascrizione di lezioni di matematica, Acts of MoodleMoot Italia 2020, p. 23-27, 2021.
  • Matteo Amore, Tiziana Armano, Cristian Bernareggi, Anna Capietto, Sandro Coriasco, Roberta Crespan, Mattia Ducci, Maria Luisa Gabrielli, Maria Francesca Guadalupi, Alessandro Mazzei, Antonio Mazzei, Adriano Sofia, Francesco Tarasconi, Riconoscimento vocale di formule, Proceedings Conference Garr 2021 Sostenibile/Digitale Selected papers, 2021, p. 35-40.