COMMPLEMENTS

Academic and Other Studies

Russell, Martin et al; "Applications of Automatic Speech Recognition to Speech and Language Development in Young Children"; ISCA-Archive, October 3-6, 1996, isca-speech.org/archive, accessed March, 2012
     The paper looks at the potential for ASR to help children develop better language skills and improved literacy. Shows that linguistic models (of the past?) are not kid-friendly. Development in this area began in 1994 with a prototype of the Speech Training Aid. Authors concluded that there are "benefits to be gained fromm deploying current speech recognition technology in tools for speech and language development in young children." "Teachers ... are now enthusiastic and believe that it is of educational benefit."

Hastings, Donald et al; "Notetaking for Deaf and Hard of Hearing Students"; c 1997
     The authors describe the importance of and proper execution of notetaking in classrooms. "... the notetaker records as much as possible of what is said in class..." including multimedia presentations. The notes should be a comprehensive collection of everything spoken in the class.

Dupont, Stephane and Luettin, Juergen; "Audio-Visual Speech Modeling for Continuous Speech Recognition"; IEEE Transactions On Multimedia, Vol. 2, No. 3, September, 2000
     The performance of current ASR systems is far from the performance achieved by humans; error rates are often one order of magnitude apart.

Srinivasan, Savitha and Brown, Eric; "Is Speech Recognition Becoming Mainstream?"; IEEE, April, 2002
     The authors, both scientists with IBM Research, look at speech recognition applications, both for dictation and transactional applications. Individuals with orthopedic challenges have found the technology invaluable.

Roberts, Kelly D.; "Voice Recognition software as a Compensatory Strategy for Postsecondary Students with Learning Disabilities"; IEEE,, Proceedings of the International Conference on Computers in Education (ICCE '02), 2002
     The author describes the low employment rate for persons with disabilities, and how assistive technology, in particular voice recognition technology for individuals with learning disabilities. Roberts finds that VRT is not effective for all persons with a LD, but success depends on attitudes and training.

Bain, Keith et al; "Speech Recognition in University Classrooms: Liberated Learning Project"; ACM, 2002, Assets 2002, July 8-10, 2002 Edinburgh, scotland
     This paper describes the use of speech recognition technology in classrooms at Saint Mary's University (Halifax, Canada). The professors trained the software, displayed the text in front of the class, and provided the transcript to students for further review. The results "were enlightening." Students were able to process both audible and verbal information, which helped integrate the information. The study was proclaimed a success, in spite of the low recognition rate of only 80%

Gutierrez, Antonio et al; "Impact of PACS and Voice-Recognition Reporting on the Education of Radiology Residents"; Journal of Digital Imaging, Vol. 18, No. 2 (June), 2005: pp 100-108
     The paper reports on the use of PACS and VRT in the training of radiology students. The authors found that proper training on the use of VRT, and acceptance for the use of the technology are important. The technology improves productivity and reduces costs; as the software has improved, recognition rates have increased to 98.5%

Grasso, Michael et al; "The Effect of Perceptual Structure on Multimodal Speech Recognition Interfaces";
     This paper describes a study on the combination of direct manipulation and speech interfaces, suggesting that performance and acceptance increase with a multimodal approach.

Boulianne, Gilles et al; "Real-Time Speech Recognition Captioning of Events and Meetings"; IEEE SLT 2008
     The authors describe the use of shadow speakers for multiple-speaker events, such as meetings and conferences. The shadow speakers are trained using 'in-domain' content - captions from previous events that contain a representative sample of the content to be translated. The larger the in-domain content, the better the overall accuracy of the text generated from the speech of the shadow speakers.

Zhili, Liu et al; "A Study and Application of Speech Recognition Technology in Primary and Secondary School for Deaf / Hard of Hearing Students"; START Centre, 2010
     The paper looks at the use of speech recognition technology in deaf/HoH education. The authors claim the technology can replace Chinese sign language for delivering information. In their study, they found 81% of students thought the text generated by ViaScribe would help recall information, provide missed information and was helpful to study. Teachers felt it was convenient for one-to-one teaching, and one teacher added that it was a good tool for providing feedback on their teaching. The authors conclude that recognition accuracy is important for acceptance, and believe the use of the technology will persist.

Hoyt, MD, Robert and Yoshihashi, MD, Ann; "Lessons Learned from Implmentation of Voice Recognition for Documentation in the Military Electronic Health Record System"; Perspectives in Health Information Management, Winter 2010
     The paper reports on a study of the use of the technology for documenting outpatient encounters. A favorable percentage of users found the technology was "accurate, faster then typing, improved note quality" and closed patent charts on the day of the encounter.

Harris, Stephen; "Research points to more advanced voice recognition"; The Engineer, November 25, 2011, theengineer.co.uk, accessed on 12/5/2011
     New research could lead to voice recognition systems that can teach themselves to better understand speakers above background noise and with different accents. scientists at Cambridge, Edinburgh and Sheffield universities are working on the improvements.

Marshall, Patrick; "Speech Recognition software now faster, more accurate than ever"; The Seattle Times, March 23, 2012, seattletimes.nwsource.com, accessed on 3/29/2012
     The "product has gotten fast and accurate enough to count on it regularly for transcriptions and dictations."