Problems of speech recognition software

There are some partial solutions, such as training the software in the speech patterns of a single user and using dynamic time-warping algorithms to match the speech to the database of samples, but they do not solve all the problems. The most complex of the speech recognition problems is identifying the context of the words being spoken.

Computer software is unable to identify the intended meaning of a collection of words, leading to a number of problems with the transcribed text.

Words that have a similar sound, such as "their" and "there", can only be accurately spelled when the context of usage is known.

For this same reason, accurate punctuation is nearly impossible for the software to place based solely on knowing the sequence of words. Click the button below to download your copy, or contact us for more information!

Content was originally published on January 20, Content was refreshed on July 9, The Trouble With Voice Recognition Software in Medical Transcription Voice recognition software VRS for medical transcription, for instance, has to make a number of arbitrary interpretations that artificial intelligence AI may not yet be fully capable to tackle perfectly.

Inability to spot errors VRS for medical transcription cannot easily spot human errors such as inconsistencies in context that human can easily detect. The type of speech needed for accurate results Not only can VRS have a negative impact on your documents , it can also become an issue for the physician. Contact Info. Sales Line: Quick Links. Client Login. All Rights Reserved.

Pinckney Marketing. Managed Hosting by Rational Pivot. Abstract Commercial speech recognition products are being used increasingly as alternate input devices for computers, particularly by persons with physical disabilities.

Publication types Research Support, Non-U. Your email address will not be published. Skip to content Home. Health Data. Shortcomings in Speech-Recognition Technology. December 15, at am 0. Real-World Consequences What follows is an example of the consequences that can occur when errors created in a medical report are overlooked.

Disclaimers Subsequent to this case, physicians and other healthcare providers began using disclaimers at the end of their transcribed medical reports to mitigate any responsibility on their part. Table 1 below provides a few examples: Table 1: Errors made by speech-recognition software Provider: The patient had influenza A.

June 18, Matthews, Kayla. December 20, Zhou, Li, Suzanne V. Blackley, and Leigh Kowalski. Stephenson, Correy. January 1, Crumbie, Joan. December 17, Sims, Lea. Horty Springer. Question of the week. March 31, Available at: www. Leave a comment.