AT to watch: Speech recognition

3D chat bubblesRecently we’ve been using and recommending speech recognition technology as a viable option for students who need alternative access to writing. We’ve also been reflecting on the use of speech recognition technology for this student population, and we think we’ve reached a critical point in terms of effective use of speech recognition for students with learning disabilities.

Many studies have indicated that speech recognition is not a viable option in the school setting-as it was only deemed appropriate for students with a specific learning profile, and there were difficulties with implementation in a school setting. While many of these arguments still hold true, two things have changed: there have been advances in the technology itself, and schools and families have started to look at a process for implementation of this type of technology.

Speech recognition technology

Speech recognition software has improved significantly. Dragon NaturallySpeaking, the leader in this category of technology, is more accurate and quite easy to use, straight out of the box. We’ve also been using Apple built-in dictation with great success. Apps have enabled speech recognition to become more universal, and more accessible. With Siri integrated into many of the apps (the microphone icon on the keyboard), students can now dictate into specific apps like iWordQ and Inspiration Maps. Notetaking apps have been transformed with the ability to use speech recognition technology to dictate notes rather than write. Check out Paperport Notes (with Dragon Speech recognition).

Process for implementation

However, the one thing that really provided us with a rationale to use speech recognition was identifying and implementing a process to teach speech recognition for our students. Last January at ATIA, we attended “Speech Recognition for Writing: A Complete Guide,” a presentation by Kelley Key and Daniel Cochrane (school-based AT specialists from Illinois). They talked about teaching a process for speech recognition-which in our opinion, is the missing link behind application of speech recognition in the school system. They discussed assessment of speech recognition, an implementation plan, and the integration of speech recognition with writing strategies. We’re currently using many of their guidelines, and finding success with many of the processes that they developed.

Another issue with speech recognition was always the training aspect in most software. Many students with reading disabilities would have difficulty with the training element of using this technology, in which the student reads passages of text to improve the software’s ability to understand the student’s voice profile. However, this was always quite difficult for students who were unable to read the text, and the result was inaccurate transcription of the dictation. For a specialized approach, check out SET-BC’s “Dragon Whispering” presentation. And good news: new speech recognition mobile technologies are quite accurate and do not require a voice profile setup.

Here are some additional resources on speech recognition:

Six ways students can use Siri

Speech recognition with Windows 8

Strategies for teachers (Alberta, Canada)

Information for parents by Reading Rockets

We’d love to hear from teachers, parents and students on how you’re using speech recognition technology in the classroom and at home. Comments appreciated!


No comments yet.

Leave a Reply