Consortium projects focus on five key challenges that represent barriers to speech recognition becoming more universally available as a solution to increase accessibility:

  1. Accuracy – reducing word error rates for speech recognition generated transcription/captioning
  2. Usability – improving overall ease of use of speech recognition in educational environments
  3. Transferability – facilitating transfer of best practices/technologies to new applications, new environments, new language settings
  4. Scalability – moving from small scale pilot to broader institutional usage
  5. Sustainability – ensuring availability of resources to support collaboration, research, innovation

To meet these challenges, the Consortium develops a yearly set of priorities and manages a diverse project portfolio.  Projects typically involve multiple Consortium partners who fill different roles as lead organization or collaborators. Resulting innovations, best practices, and new technologies are shared with the entire Consortium membership.

 

2011 Projects

Speech Recognition Benchmarking
Classroom Testing
Synote

Language/Lecture Model Development
Business Case Exploration
Accessible OCW / MOOCs

Past Projects